Geochemistry and petrology of the Loon Lake pluton, Ontario

Geochemistry and Petrology of the Loon Lake Pluton, Ontario JAROSLAV DOSTAL Department of Geology, Dalhousie University, Halifax, Nova Scotia B3H3J5 Received October 23,1974 Revision accepted for publication April 7, 1975 The Precambrian Loon Lake pluton, Ontario, consists of two main zones-acore of monzonite with a rim of younger quartz monzonite. Several isolated bodies of older diorite and syenodiorite occur within the pluton. The variations in the chemical and mineralogical composition of diorites and syenodiorites are due to both magmatic differentiation and hybridization. The trends of the variations of major elements, Rb, T1, Sr, Ba, and rare earth elements in monzonite are consistent with fractional crystallization mainly of feldspars; this fractionation probably involved flowage differentiation. Fractional crystallization and contamination of monzonitic magma by anatectic granitic melt probably played adominant role in the genesis of quartz monzonite. Monzonite and quartz monzonite are believed to have formed from a magma of lower crustal or upper mantle origin. While part of this magma intruded as monzonite, another part which evolved further generated quartz monzonite. Le pluton precambrien de Loon Lake, Ontario, comprend deux zones: un coeur de monzonite avec une couronne plus jeune de monzonite quartzique. A I'interieur du pluton on retrouve de petites masses isoldes plus anciennes de diorite et de syenodiorite. La differenciation magmatique et I'hybridation expliquent les variations de composition chimiqueset mineralogiques des diorites et des syenodiorites. La distribution des elements majeurs, des dldments traces Rb, TI, Sr, Ba, et des T.R. dans les monzonites est en accord avec un processus de cristallisation fractionnee principalement des feldspaths; cette cristallisation implique probablement une differenciation mecanique. La cristallisation fractionnee et la contamination des magmas monzonitiques par un liquide anatectique de composition granitique, ont probablement joud un r6le dominant dans la genese des monzonites quartziques. Monzonites et monzonites quartziques auraient dtd formees par un magma provenant de la crofite inferieure ou du manteau superieur. Pendant qu'une partie du magma formait les monzonites, une autre partie qui evoluait inddpendamment allait former les monzonites quartziques. [Traduit par le journal] 1 Introduction The southwestern part of the Grenville Structural Province of the Canadian Shield contains many small isolated Precambrian plutons predominantly of monzonitic or syenitic composition. The origin of these plutons has been a matter of controversy. Whereas Wynne-Edwards (1957, 1967) suggested that replacement processes play an important role in the formation of these rocks, Currie and Ermanovics (1971) have advocated that the plutons originated by anatectic melting in situ or nearly so, followed by desilication by vapor diffusion. Remobilization of the underlying rocks of the Grenville basement (e.g., Sauerbrai 1966; Ermanovics 1967) and magmatic differentiation (Lumbers 1967) have also been advanced to explain the origin of the rocks of these plutons. The disputed origin of these bodies is obviously another case of the 'syenite problem'. The purpose of this paper is: (1) to present some geochemical and petrological data related Can. J. Earth Sci., 12, 1331-1345 (1975) to the Loon Lake pluton, which is representative of a number of similar monzonitic plutons found in this part of the Canadian Shield; (2) to attempt to evaluate the origin and evolution of this body; and (3) to contribute to the understanding of the origin of syenitic rocks in general. Geological Setting The Precambrian Loon Lake pluton is situated in Chandos Township, Peterborough County, southern Ontario. It is a roughly ovoidal-shaped body, about 40 sq. km in area, outcropping within the metasedimentary-metavolcanic rocks of the Chandos Subgroup (Shaw 1972) in the Grenville Structural Province of the Canadian Shield. Near the pluton, the Chandos Subgroup is composed, in order of decreasing abundances, of gneisses, marbles, and amphibolites. This sequence underwent complex Proterozoic metamorphic events. The main progressive regional metamorphism, which culminated at about 1332 CAN. J. EARTH SCI. VOL. 12, 1975 FIG.1 . Generalized geological map of the Loon Lake pluton region. + 1125 25 m.y. ago (U/Pb dates on zirconSilver and Lumbers 1965) is of Miyashiro's (1961) "low pressure-intermediate type" in this area (Chesworth 1967, 1971). More specifically, the metamorphic grade of the Chandos Subgroup corresponds to amphibolite facies and lies above the first sillimanite isograd (Shaw 1962; Lumbers 1967). On these already regionally metamorphosed rocks, the intrusion of the Loon Lake pluton superimposed a contact metamorphism of the pyroxene-hornfels facies (Shaw 1962; Chiang 1965). The contact metamorphism, which took place at a greater depth than a classical shallow-seated contact metamorphism, also produced partial melting of some country rocks (Dostal 1975). General Geology of the Pluton The Loon Lake pluton (Chandos Lake pluton of Saha 1959) is steeply dipping and has the shape of an asymmetrical funnel (Cloos 1934; Saha 1959; Shaw 1962). The pluton, which is structurally independent of the regional pattern, was forcefully injected; it shouldered aside and in places refolded the walls (Shaw 1962). It is a zoned composite body, consisting of two main zones which have a roughly concentrical arrangement (Fig. 1). The core of the pluton is formed of monzonite, while the outer zone is composed mainly of quartz monzonite. Several elongated bodies of more basic rocks (syenodiorite, diorite) up to 500 m long occur as isolated masses within the pluton, most of them in the central portion. Apart from these three rock types, a structurally conformable septum or screen of partially assimilated and migmatized biotite gneisses occurs in the outer quartz monzonitic zone of the pluton (Fig. 1). The pluton appears to represent multiple intrusions of magma. In order of decreasing age DOSTAL: THE LOO1V LAKE PLUTON the rock types are: basic rocks, monzonite, and quartz monzonite (Saha 1959; Shaw 1962). On geological and petrographical grounds, Saha (1959) and Shaw (1962) argued that the basic rocks are inclusions or roof pendants not necessarily genetically related to the bulk of the pluton. The contact between the monzonite core and the outer quartz monzonite zone generally appears to be gradational over a short distance. In places, however, quartz monzonitic rocks of the outer zone also cross-cut the monzonite (Saha 1959). According to Saha (1959) the older monzonitic intrusion was probably not yet completely consolidated and still behaved somewhat plastically as the quartz monzonite was intruded. Foliation, which is present mainly in quartz monzonite, was interpreted by both Saha (1959) and Shaw (1962) as primary flow foliation. The extensive mobilization of hornfelses in the contact aureole and some textural features (Dostal 1973) indicate that the pluton intruded already heated, regionally metamorphosed rocks and diastrophism was probably still continuing when the intrusion took place. This suggests that although the pluton was emplaced well after the culmination of regional metamorphism (Saha 1959; Shaw 1962), dated at 1125 25 m.y., both these events might be part of one major cycle of orogenesis. The pluton, however, is older than the intrusions of pegmatites and calcite-fluorite-apatite-bearingveins, which were emplaced at about 1000 m.y. ago (Silver and Lumbers 1965; Sr, Pb, and Ar methodsShafiqullah et al. 1973). The estimated age of the pluton of about 1075 f 75 m.y. is comparable with those of monzonitic plutons in the nearby Westport area, Ontario, which might be of the same or similar origin. + Petrography of the Pluton Quartz monzonites, which make up the outer zone of the pluton, are medium- to coarsegrained, with biotite as the predominant mafic minera1.Varieties containing subordinate amounts of hornblende in addition to biotite occur rarely. Plagioclase has a composition ranging from An,, to An,. Accessory amounts of magnetite, sphene, apatite, zircon, allanite, and secondary calcite, epidote, and chlorite are present. The texture is hypidiomorphic granular, with features of cataclastic and mortar texture in places. 1333 Monzonites form the core of the pluton. In a few cases, however, they also occur in the outer quartz monzonitic zone. They are coarse-grained and leucocratic, with 2-10% of mafics. The principal ferromagnesian mineral is biotite, but in some specimens, hornblende or, rarely, clinopyroxene are also abundant. The composition of plagioclase varies from about An,, to An,, . Accessory sphene, magnetite, apatite, and secondary calcite, epidote, chlorite, and muscovite are frequently encountered. The texture of these rocks is allotriomorphic to hypidiomorphic granular, with locally pronounced features of mortar and cataclastic texture. The monzonitic core of the pluton also displays subtle but systematic progressive acidification towards the margin (Saha 1959; Dostal 1973). The anorthite content of plagioclase and the content of mafic minerals decrease from the center toward the margin. These trends together with the presence of clinopyroxene, the first mafic mineral to crystallize, only at the center of the monzonite zone, indicate that the rocks occurring in the center of the monzonite core started to crystallize earlier than those around the margin. The variations of mineral composition were probably present in the partly crystalline monzonite magma before its emplacement; Shaw (1962) suggested that when the magma was intruded it was crystalline to a notable extent. This indicates that the systematic variations in the monzonite core might have been due to flowage differentiation where the early precipitated crystals moved from the wall towards the center causing this concentric variation. Syenodiorite and diorite that form the basic inclusions imperceptibly pass into each other. Petrographically, syenodiorite differs from diorite mainly by higher contents of K-feldspar and biotite. The basic rocks usually have hypidiomorphic granular to cataclastic texture and are composed chiefly of plagioclase (andesine), hornblende, and biotite, with varying amounts of clinopyroxene, quartz, and K-feldspar. Calcite, magnetite, sphene, and apatite are present in accessory amounts. The large scaly aggregates of biotite and patches of K-feldspar replacing other minerals appear to be related to the hybridization of basic rocks by younger felsic magma (Saha 1959, Shaw 1962). The contents of these 'later' minerals decrease inward from the edges of the basic inclusions. These rocks probably were originally gabbro-diorites (Saha 1959). TABLE 1. Chemical analyses of the rocks from the Loon Lake pluton SiOl Ti02 A1203 Fe20s FeO MnO MgO CaO Na20 K20 pzos H20 Hz0 co2 TOTAL D.I. Rb TI Sr Ba La Ce Nd Sm Eu Gd Tb Ho Yb Lu NOTE:Oxides in wt %; trace elements in p.p.m. except T1, which is given in p.p.b.; n.d. = not detected The lacations of all samples together with their modal compositions are given in Dostal (1973). I. diorjte (321) 8. hornblende-biotite monzonite (96) 2. dior~te(199-5) 9. biotite quartz monzonite (26) 3. syenodiorire 153-5) 10.biotite quartz monzonite (1 15) 4. hornblende-b~outcmonzonltc I20n 11. biotite ouartz monzonrte 168) 5. cl~nopbro~ent-hornblende-biotite monzonlte (198) 12. biotite quartz monzonite (27) 0, hornblende-b~ot~te rnonzonlte (228) 13. biotite quartz rnonzonite (253) 7. biotite monronite (251) DOSTAL: T H E LOON LAKE PLUTON Geochemistry of the Pluton Samples of 9 diorites and syenodiorites, 13 monzonites, and 12 quartz monzonites were analyzed for major elements and some of them were also analyzed for trace elements. Some of the results are presented in Table 1. Complete data together with the analyses of the mineral phases may be obtained on request from the author. Major elements and Rb, Sr, and Ba were determined by standard rapid methods adapted in the Department of Geology, McMaster University, Hamilton, Ontario. TI was analyzed by atomic absorbtion spectrometry using the method of Fratta (1974), while rareearth elements (REE) were determined by a radiochemical neutron activation technique adapted from that of Denechaud et al. (1970). The procedure involves post-irradiation chemical group separation of REE and its radio-assay by y-ray spectrometry using a high resolution Ge(Li) detector and multichannel analyzer. The accuracy and precision of these analyses are given in Fratta and Shaw (1974) and Dostal (1975). Major Elements A common problem of many composite plutons containing several intrusive phases is whether such intrusions have been formed from a single parental magma and if so, what is the nature of the original magma. In order to evaluate whether the behavior of the felsic rocks of the pluton is consistent with some process of magmatic differentiation, their compositions have been plotted in the projections of the Or-Ab-An-Q-H,O system, because monzonitic and quartz monzonitic rocks contain more than 80% of normative feldspars and quartz. Figure 2 shows the normative composition of the felsic rocks from the pluton in relation to the Q-Ab-Or ternary diagram. Most of the quartz monzonitic rocks fall within Winkler and Von Platen's (1961) granitic field, while the rest of the quartz monzonites fall into the low temperature trough. A comparison with the experimentally determined melting minima shows that the monzonites plot toward the Or corner from the alkali feldspar minima. However, this shift is probably typical of syenitic rocks in general and was attributed to the confluence of other phases, notably calcic pyroxene (Morse 1968). Figure 3 shows the position of the felsic rocks from the pluton in the Or-Ab-An diagram and their FIG.2. Normative composition of felsic rocks from the Loon Lake pluton in relation to the Q-Ab-Or ternary projection. The solid line encloses analyses of 1190 granitic rocks (Winkler and Von Platen 1961). Dasheddotted and dotted lines represent field boundaries at indicated Ab/An ratios at 2 kb PHZO.Squares denote ternary minima or eutectics at various PHlo with an Ab/An ratio of 2.9 (W I = 2 kb; W 2 = 4 kb; W3 = 7 kb; W 4 = I0 kb PH20)(both taken from Winkler 1967). 5 marks the position of the temperature minimum at 2 kb and Ab/An = oo (Tuttle and Bowen 1958). 0 = monzonite; = quartz monzonite. FIG.3. Normative composition of felsic rocks in relation to the An-Ab-Or ternary projection. The solid lines are the boundaries of the low temperature trough which range from Or29Ab71at 10 kb t o Or45Ab55 at 1 kbPH,,,. Dashed lines show uncertainty due to the possibility of analytical error (Kleeman 1965). The irregular boundary is the 2% contour of Tuttle and Bowen (1958, Fig. 67) for granitic rocks that contain more than 80% normative Ab Or Q. Symbols are same as on Fig. 2. + + f 336 CAN. J. EARTH SCI. VOL. Wt% 12, 1975 Wt % 10 70 a *. . .. . . . ......... . -.. m m 15 - I ' u m 08 -I - . .. . a 5 - *,n . . a*: - , _t* lo-5.. I 20 - *I2O3 -: r 1 - 00 8 2 Coo , Ki A - 05 - I2 - . ' . 9 I1 -4 4 + - . . .. . - . 5- No20 oe 05- @@ LO* - : ID - : %* 5 . : .. . . : ; : , : : . - - . X 2-. I-., 0.9 - 08 2rX 0,6 5 a ' B b ' i ' & ' k ' l I K ) D. I. . . . .. ..... e C ZFe 06 FeO ' m ' m I 1 . 75 . *8 01 -. a ' I - 2 , , . . . . . . . . : . : w . . 0 3 0 ' & ' j , ' $ , ' , ' , ' w ' l m D I FIG.4. Variation of oxides of major elements as a function of differentiation index. W = basic rocks; 0 = monzonites; = quartz monzonites; x = leucogranite 259. relation to the low temperature trough of Kleeman (1965). Most of the quartz monzonites and monzonites fall into the thermal trough, and the rest lie on the plagioclase side of the low temperature region within the 2% contour of Tuttle and Bowen (1958) for 'normal granites'. The position of the felsic rocks in the Or-AbAn-Q-H,O system suggests that monzonite and quartz monzonite probably resulted from "crystal-liquid equilibrium" (Tuttle and Bowen 1958). The contents of major elements in the rocks from the pluton have been plotted on the variation diagrams against the differentiation index (D.I.) of Thornton and Tuttle (1960) in Fig. 4. The lower contents of Al,O,, K,O, and Na,O and higher content of SiO, in quartz monzonites in comparison with monzonites of comparable D.I. values reflect the higher content of modal quartz and lower amounts of modal feldspars in the former rocks, while the contents of mafic minerals remain roughly constant. Considering the three main rock types as a series, Fe, Ti, Ca, and Mg decrease toward the higher D.I. values. The diagram shows a sharp increase of the Fe/(Fe + Mg) ratio in monzonites and quartz monzonites with the decrease of their element concentrations. The basic rocks, however, display only relatively small variations of this ratio 1 DOSTAL: THE LOON LAKE PLUTON in spite of the large variations of the absolute concentrations of these elements. When monzonitic and basic rocks are treated as a series on these variation diagrams, most of them show smooth trends from the basic to the acid end with nearly complete chemical gradation between them. Some of the observed variations of transition rocks appear to be approximately 'straight line' trends not readily consistent with magmatic differentiation. Such trends, however, might have been formed from two magmas with intermediate types being developed by assimilation and hybridization (Nockolds and Allen 1956). But the fact that at least some of the variations in the chemical composition of basic rocks are of primary magmatic origin is indicated by the variation of the Fe/(Fe Mg) ratio or the variation of modal basicity of plagioclase. On the other hand, some variations in basic rocks are probably of secondary origin, as suggested by petrographic features (e.g., late microcline and biotite, textural evidence for replacement in basic rocks), which indicate an interaction between basic and felsic rocks and magmas. The process of hybridization is also consistent with the remarkably high content of K in some basic rocks. Thus it is difficult to evaluate to what degree the smooth trends in these diagrams are due to primary magmatic variation and to what degree they represent hybridization or assimilation. The variation trends for felsic rocks are compatible with those produced by magmatic differentiation and suggest a genetic relation between quartz monzonites and monzonites. It appears that the main difference between them is the quartz content. It is also of interest that leucogranites (e.g. 259), which are thought to be the product of partial melting (Dostal 1975), represent the 'most differentiated' rocks on the variation diagrams. 2000 1000.. Sr 500 - 200 - 1337 . loo- + Rb, TI, Sr, and Ba The general increase of Rb from basic toward more acid rocks (Fig. 5) is similar to trends reported from similar suites of rocks elsewhere and predicted by crystal-chemical principles. Rb concentrations in monzonites and some quartz monzonites are, however, notably lower than the world averages calculated for comparable rocks (Heier and Adams 1964). This depletion of Rb is also reflected in the K/Rb ratios. Relative to the 'normal' K/Rb values of FIG.5. Variation of Rb, Sr, Ba, and La as a function of differentiation index. Symbols are same as on Fig. 4. crustal rocks (=230), the ratios in the rocks of the Loon Lake pluton are very high, particularly in the monzonites and some quartz monzonites (- 600-1 100). The constituent minerals of monzonites also have high K/Rb ratios. The K/Rb ratios of biotite ( E 200) are significantly higher than those of typical granitic and metamorphic biotites 1338 CAN. J. EARTH S( :I. VOL. 12, 1975 FIG.6. Variation of the K/Rb ratio as a function of Rb in the rocks of the pluton. Symbols are same as on Fig. 4. (Lange et al. 1966; White 1966; Whitney 1969). Likewise, K-feldspars from monzonites have very low contents of Rb (1.150 p.p.m.), resulting in an abnormally high K/Rb ratio. The "fractionation coefficient" FIB = (K/Rb in K-feldspar)/(K/Rb in biotite) (Lange et al. 1966) ranges in the felsic rocks of the pluton from 3.4 to 4.4, and has a mean value of 3.8 (Dostal 1973). However, comparable values from these mineral pairs have been reported from granitic rocks which have 'normal crustal' values of the K/Rb ratio, and also from synthetic sanidine-phlogopite pairs (Beswick and Eugster 1968; Beswick 1973). The rocks from the Loon Lake pluton are shown on a plot of Rb us. K/Rb in Fig. 6. They are separated into two distinct trends, similar to those of well defined magmatic sequences (cf. Dupuy 1970) and attributed to magmatic differentiation. The first field, with lower Rb values, includes the basic rocks (in this case, basic rocks include only rocks which petrographically appear to be least modified), while. the other represents monzonites and quartz monzonites. The close geochemical coherence between Rb and T1 is also apparent in the rocks studied. The values of the Rb/Tl ratios in rocks from the pluton (Fig. 7) are relatively close to the crustal average for this ratio (1.150) given by Alburquerque and Shaw (1972). On the other hand, the KIT1 ratios (Fig. 7) particularly for monzonites are very high in comparison with the average value for crustal igneous rocks, 30,000 (Albuquerque and Shaw 1972), indicating a strong depletion of T1 relative to K in these rocks. The impoverishment of TI is of a magnitude similar -- FIG.7. Relations between Rb and TI (top) and between K and TI (below) in the rocks from the pluton. The ratio of Rb/TI = 150 and those of K/TI = 10 x lo5 and KIT1 = 3 x lo5 are shown. Symbols are same as on Fig. 4. to that of Rb, as suggested by the 'normal' Rb/Tl ratio. The high K/Rb and KIT1 ratios in monzonites and some quartz monzonites are of considerable petrogenetic interest. The similarity of the "fractionation coefficient" FIB from the Loon Lake pluton to those from other granitic rocks, and also the high K/Rb ratio in other monzonite and nepheline syenitic bodies in this region (Payne 1966; Sauerbrei 1966), the lack of any apparent relationship between modal biotite and the K/Rb ratio of the host rocks (Dostal 1973), and the high K/Rb ratio not only of feldspars but also of biotite suggest that the high K/Rb ratio in monzonites is probably inherited from the original liquid from which they crystallized. The variation of Sr in the rocks from the pluton as a function of D.I. is shown in Fig. 5. A similar trend, a decline toward the DOSTAL: THE LOON LAKE PLUTON FIG.8. Rare-earth element distribution in basic rocks from the Loon Lake pluton. 0 = dio= diorite 199-5; A = syenodiorite 53-5; 0 = hornblende from diorite 70; W = diorite 321; rite 70; and A = biotite from diorite 70. more acid rocks, is displayed by Ba (Fig. 5). These trends for Sr and Ba are also reflected by the steep variations of the Ca/Sr and K/Ba ratios (not shown), which are characteristic of the late stages of highly differentiated magma (Nockolds and Allen 1953, 1954). Since Sr and Ba are predominantly concentrated in feldspars, their smooth variation trends in the felsic rocks of the pluton are consistent with the crystallization of feldspars as the major phases, which would preferentially take up the Sr and Ba from the liquid so that the progressively more residual rocks would be depleted in these two elements. Rare Earth Elements (REE) In order to correlate the REE abundances with the petrochemistry of the rocks from the pluton, the La concentrations have been plotted against D.I. in Fig. 5. The La content in basic rocks shows a tendency to increase with the increase of the D.I. values. Some basic rocks, however, have a concentration of La very similar to that of basic monzonites despite the fact that their D.I. values are markedly lower. With regard to monzonites, the large increase of La toward the more acid rocks indicates that monzonites underwent extensive magmatic differentiaton. Quartz monzonites, however, lie well off the variation trend of monzonites. The evaluation of REE variations may be complicated by the fact that the bulk of REE in quartz mon- zonites is present in accessory minerals (Dostal 1973). The REE distributions in the rocks from the Loon Lake pluton are given in Figs. 8 and 9 conventially normalized to the average chondrites (Frey et al. 1968). The REE patterns of basic rocks (Fig. 8), which are well-fractionated with gradual enrichment from Lu to La, remain substantially unchanged, despite the large variations of the absolute REE concentration. One sample (53-5) shows a positive Eu anomaly indicating the incorporation of an excess of feldspars. If the Eu anomaly was generated during magmatic differentiation, it would suggest that this rock is at least partly of cumulative origin. The positive Eu anomaly is, however, also consistent with the process of hybridization. An addition of feldspars during hybridization could produce the observed Eu anomaly; this rock contains in mode more than 20% of Kfeldspar which appears to be of late origin. Although the variations of REE in basic rocks in general are suggestive of magmatic differentiation, they appear to be also compatible with hybridization or assimilation. Since monzonites (apart from those which have low D.I. values) have a higher absolute content of REE than basic rocks, then their partial hybridization by monzonite might produce the observed trends. It is of interest that biotite formed by the replacement of hornblende in these rocks has a REE pattern very similar to that of hornblende CAN. J. EARTH SCI. VOL. 12, 1975 1000 . FIG.9. Rare-earth element distribution in felsic rocks from the Loon Lake pluton. Top: 0 = monzonite 207; = monzonite 198; = monzonite 228; = monzonite 251; and A = mon= quartz monzonite 115; @ = quartz monzonite 27; and A = quartz zonite 96. Below: rnonzonite 26. (Fig. 8), suggesting that the REE distribution did not change significantly during this alteration. The REE patterns of monzonites (Fig. 9) are well fractionated and progressively enriched from Lu to La relative to chondrites (except for Eu). The fractionation patterns are similar but differ slightly in absolute REE contents and in degree of relative fractionation. More variation is seen for Eu and the Eu/Eu* values (which indicate the ratio of the observed europium abundance-Eu-to that predicted-Eu*-by graphical extrapolation between the values of Sm and Tb or Gd). Figure 9 shows a general tendency for the La/Yb ratio to increase with the increase of the absolute REE abundance and the decrease of the Eu/Eu* ratio. These varia- a tions are compatible with the trends attributed to fractional crystallization. A tendency of the Eu/Eu* ratio of monzonites (Fig. 9) to decrease toward the more acid rocks is similar to the variation trends for Sr and Ba (Fig. 5). The parallel behavior of Eu suggests that it is probably partly present as Eu2+. The relative depletion of Eu paralleled by the fall of Sr and Ba with the increase of the D.I. values is readily explained by the removal during magmatic differentiation of significant amounts of feldspars which concentrate not only Sr and Ba, but also EU". In general, the REE distribution patterns of quartz monzonites (Fig. 9) are similar to those observed in other granitic rocks (Herrmann 1970). They have the typical enrichment in light I I I DOSTAL: THE L.OON LAKE PLUTON REE relative to chondrites, with or without a small Eu depletion and with relatively small fractionation of heavy REE. But the REE patterns of quartz monzonites differ from those of acid monzonites (i.e., monzonites with high D.I. values). The former have less fractionated patterns, lower absolute contents of REE and lack significant negative Eu anomalies. These differences once again indicate that some quartz monzonites are less fractionated than acid monzonites. Petrogenesis of the Basic Rocks Despite the fact that the variation trends between the basic and monzonitic rocks were affected by hybridization, it may be suggested that the basic rocks could originally be genetically related to the bulk of the pluton. But if the chemical trends of these basic rocks, which petrographically appear to be the least altered, do not reflect only secondary changes, then the variations of some elements and element ratios might argue against a genetic relation between the basic and monzonitic rocks. The distinctly separated trends for basic and felsic rocks in plots such as K/Rb us. Rb might suggest that these two rock types do not belong to a comagmatic series and that they were not produced by magmatic differentiation of a single magma. Petrogenesis of the Monzonitic Rocks The monzonitic rocks as a series show a large and systematic variation in their chemistry. These variations are consistent with extensive fractional crystallization. Such a process probably involved the crystallization of large amounts of feldspars with K/Rb, K/TI, Eu/Eu* ratios, and concentrations of Ba and Sr higher than the melt and with Ca/Sr, Rb/Sr, K/Ba ratios, and contents of Rb and T1 lower than the melt. The crystallization of feldspars was also likely accompanied by smaller amounts of mafic and possibly accessory minerals (sphene and probably also apatite) with La/Yb and Fe/ (Fe + Mg) ratios lower than in the melt. Thus the residual liquid from which petrogressively more acid monzonite crystallized was accordingly depleted in Ba, Sr, Eu, and heavy REE and enriched in Rb, TI, total rare-earths, and a relative concentration of light REE. The high concentrations of Ba, Sr, the unusually high K/Rb ratio, and the positive Eu anomaly in the 1341 basic, least fractionated monzonite (with the lowest D.I. values) emphasize the important role of feldspar fractionation, and these features may be attributed largely to feldspar accumulation. But monzonites as a whole are not the product of feldspar accumulation and probably crystallized from monzonitic magma. The negative Eu anomalies in acid monzonites, the petrography, the position in the Q-Ab-Or projection, and the chemical variation trends, indicate that feldspars are not in excess. With regard to the actual mechanism, it seems significant that the monzonites with low D.I. values and La contents, positive Eu anomalies and high Ba and Sr concentrations occur in the center of the monzonitic core of the pluton. Such a distribution is in accordance with a process of flowage differentiation, where earlyformed crystals (mainly feldspars) are concentrated toward the core, and thus they could be partly of a cumulative nature while progressively more differentiated liquids would be concentrated at the outer zones. Several possible hypotheses for the origin of monzonitic magma of the Loon Lake pluton merit evaluation: (1) contamination of a basic magma; (2) contamination of an acid magma; (3) partial melting of crustal rocks; (4) derivation from a more basic magma. (I) Contamination of the Basic Magma The origin of syenitic and monzonitic rocks due to an assimilation of crustal rocks by basic magma is not very compelling for the genesis of monzonite from the Loon Lake pluton. The assimilated material would have to be enriched in alkalies, particularly in potassium, far above the abundances in common rocks. Another problem involving this hypothesis is the relative lack of intermediate rocks. (2) Contamination of Acid Magma The alternative hypothesis of assimilation is that of Fenton and Faure (1969), who invoked the process of assimilation of gneisses and marbles by granitic magma for the origin of the syenites. Low concentrations of Ca, Mg, and Fe in monzonites, however, negate large-scale assimilation. This assimilation would also probably not explain the high content of alkalies in monzonite. (3) Partial Melting of Crustal Rocks The partial melting of quartz-poor feldspathic 1342 CAN. J. EARTH SCI. rocks in the deeper crust appears to be consistent with most of the data. But the high K/Rb ratios of monzonites produced by a simple anatectic process would require derivation from a parent that also has a high K/Rb ratio. Thus the origin of monzonite by partial melting of common crustal rocks with a 'normal' value of about 230, is not compelling. The difficulties of generating the high K/Rb ratio apply also to the partial melting of dioritic rock or amphibolite, assuming that their K/Rb ratio does not differ very significantly from the crustal value. Anatexis of these rocks would not yield a melt with an appropriate trace element concentration, with the presence of amphibole in the residue (Hart and Aldrich 1966). On the other hand, the higher degree of partial melting would probably not generate a magma with such large absolute contents of alkalies and low concentrations of Ca, Mg, and Fe. The difficulties with high absolute contents of alkalies and low concentrations of Ca, Mg, and Fe also apply to another alternative explanation-the selective melting of hornblende, which has a high K/Rb ratio. A more plausible explanation seems to be a two-stage melting process in the lower crust, perhaps similar to that proposed by Reynolds et al. (1969) and Green et al. (1972). A first partial melting stage would produce a granulitic residuum with a high K/Rb ratio, followed by a second stage with a higher degree of melting where feldspars contribute significantly to the melt formed. This subsequent melting in the lower crust would generate magma with a high K/Rb ratio which might correspond to monzonite. It is difficult, however, to reconcile this melting episode of granulite-type material with the content of some lithophile elements in monzonite. ( 4 ) Derivation from a More Basic Magma The chemical composition of monzonitic rocks is consistent with this hypothesis. Elements such as K, Ba, and REE which are strongly concentrated by a process of fractionation, are enriched in monzonites while the ferromagnesian elements (Ni, Co, V, Cr (McCammon 1968)) are depleted. Also the similarity of the chemical composition of monzonite to some trachytic rocks indicate that monzonite from the pluton may have been derived from a basic magma. VOL. 12, 1975 above, leads to some speculations with respect to the nature of the 'source material'. Payne (1966) has suggested that the high K/Rb ratios of nepheline syenite from the neighboring Blue Mountain body indicate a mantle origin for the nepheline syenite magma. Such an origin of nepheline syenite is also corroroborated by Sr isotope data (Krogh 1964; Krogh and Hurley 1968). The Loon Lake pluton displays remarkable mineralogical and chemical similarities to zoned plutons in the Westport area, Ontario (e.g., Westport and Gananoque plutons). The data of Sauerbrei (1966) show that these plutons in the Westport area also have a high K/Rb ratio (-550). In this respect it is of interest that Krogh and Hurley (1968) inferred the mantle origin for these plutons on the basis of a low initial 87Sr/86Srratio. The similarity between these bodies and the Loon Lake pluton even in their high K/Rb ratios, which are also comparable to those from the apparently mantle-derived Blue Mountain nepheline syenite, may indicate that the magma from which monzonitic rocks of the Loon Lake pluton were derived is of upper mantle or of lower crust origin. The high K/Rb ratio also suggests that monzonitic magma was not contaminated to a large degree by crustal rocks with a 'normal' K/Rb ratio, since significant contamination would probably cause a sizeable decrease in this ratio. The high contents of K, Ba, Sr, and REE, strongly fractionated REE patterns, and low concentrations of ferromagnesian elements in monzonites indicate that they may be derived from a very small degree of partial melting and/or extensive fractional crystallization of more primitive material. The felsic rocks from the pluton show chemical and mineralogical similarities to monzonites, syenites, mangerites, and quartz monzonites associated with anorthosites. They are also characterized by high K/Rb ratios (Reynolds et al. 1969; Green et al. 1972). In the Adirondack complex, these rocks even have an age (Heath and Fairbairn 1968) comparable to that of the pluton. This might suggest that the felsic rocks of the Loon Lake pluton may perhaps also be related to the anorthosite suite. Petrogenesis of Quartz Monzonites With respect to the petrogenesis of quartz monzonitic rocks, the important problem is the Source Materials of Monzonites The Rb and T1 depletion in monzonites, if it is relation between them and monzonites. The an original feature of the magma as suggested similarity and continuity of their chemical pro- DOSTAL: THE LOON LAKE PLUTON ! ,1 I 1 perties indicate that these rocks are genktically closely related, probably comagmatic. Some quartz monzonites, however, are less fractionated than acid monzonites. Consequently this suggests that these two rock types as a series are not simply products of continuous magmatic differentiation of a single magma leading to accumulation of residual quartz and that a more complex process was probably involved. Accepting a genetic relationship between monzonite and quartz monzonite, it appears that the models consistent with the greatest proportion of the observed data would incorporate basically two processes: (a) from a single magma two portions were tapped and each of these subsequently underwent further differentiation along independent but similar paths. (b) monzonitic magma was in part intruded into the present position and in part it fractionally melted and reacted with crustal rocks, producing the quartz monzonite and then subsequently intruded. The first model is comparable to that proposed by Upton (1960) for the Kungnat syenite complex in southwestern Greenland. Such a mechanism may probably explain most features of chemical variations in quartz monzonites. The variation of REE, however, does not appear to be readily compatible with this process. But the evaluation of the variation of REE is complicated by the fact that the bulk of REE in quartz monzonite is present in accessory minerals (Dostal 1973), which might mask or hinder this trend. The second model is similar to that suggested by Barker et al. (1972) for the Pikes Peak batholith, Colorado. The temperature imposed on the wall-rocks by the pluton was well above the temperature of the beginning of partial melting of at least some of them. Thus it is probable that magma which ascended through crustal rocks could hardly avoid a partial melting of gneissic or granitic rocks where it came into contact with them. Partial melting of the latter might produce a granitic anatectic melt, which by mixing with monzonite could form quartz monzonite. In this respect, it is of interest that leucogranite, probably generated by the partial melting of the Apsley gneiss (Dostal 1975), lies at the most differentiated end of the trends on the variation diagrams. It shows that contamination of monzonitic magma by anatectic melt might have produced the many observed variations of quartz 1343 monzonitic rocks. The anatectic melt was probably produced in depth, as there is geological evidence that the bulk of the quartz monzonitic zone of the pluton could not be generated in situ or nearly so. The Bancroft area, including the Hastings lowlands with the Loon Lake pluton, is underlain by a thick granitic or gneissic layer (Jacoby 1971), which probably corresponds to the basement complex of the Grenville Supergroup (Wynne-Edwards 1972). The overall chemical and petrographical similarities of the assumed rocks of the basement complex to the Apsley biotite gneiss suggest that the products of the partial melting of either of these rocks would be rather similar, particularly since the composition of the partial melts corresponds to the low melting fractions of these rocks. Although the anatectic melt might have varied from place to place, these variations were probably not large. This might suggest that the leucogranite resulting from partial melting of the Apsley gneiss could also resemble a partial melt of the basement rocks which might have mixed with monzonite. The 'mixing' process can readily explain a variation of REE in quartz monzonites. But it does not seem very likely that this process alone could produce all variation trends observed in quartz monzonite. Many of these trends are similar to those produced by fractional crystallization and strongly suggest such a process. The two models proposed are not, however, mutually exclusive. Each only stresses the relative significance of a particular mechanism, even if both processes were operative. When monzonitic magma moved upward through a deeper zone of crust, where the rocks were already undergoing high-grade metamorphism and possibly partial melting, it is likely that the magma became contaminated with anatectic granitic melt. On the other hand, if quartz monzonitic magma was already emplaced as a crystal mush (Shaw 1962), it is probable that the magma underwent a certain degree of fractional crystallization, at least during its emplacement (e.g., by flowage differentiation, crystal settling). But it appears that the process of fractional crystallization was predominant. Summary The Loon Lake pluton is similar in many aspects to mesozonal plutons of Buddington (I 959). It is a late-tectonic funnel-shaped intru- 1344 CAN. J. EARTH SCI. VOL. 12, 1975 + sion emplaced about 1075 75 m.y. ago into a region where diastrophism was still continuing. Variation in chemical and mineralogical composition of basic rocks are due to both hybridization by felsic magma and magmatic differentiation. These rocks might not be genetically related to the bulk of the pluton. The distinct chemical variations within monzonite are consistent with extensive fractional crystallization, mainly of feldspars. The actual mechanism of this fractionation probably involved flowage differentiation. Monzonite and quartz monzonite were both formed from a magma that was generated by partial melting of lower crustal/upper mantle rocks, probably followed by fractional crystallization. A part of this magma was intruded as monzonite and another more evolved part formed quartz monzonite, which was intruded following the previous intrusion of monzonite into its present position. Quartz monzonite was generated by fractional crystallization of monzonitic magma, and by 'mixing' of monzonitic magma with granitic melt produced by partial melting of crustal rocks, perhaps those of the basement complex of the Grenville Supergroup. The former process, however, probably prevailed. Acknowledgments I am indebted to Dr. D. M. Shaw for valuable discussions and continued encouragement. Mr. J. R. Muysson performed the analyses for the major elements, Rb, Sr, and Ba, while Mr. P. Fung undertook the determination of T1. Thanks are due to Drs. C. A. R. Albuquerque and W. Chesworth for critical comments on the text. I acknowledge support by research grants to D. M. Shaw from the National Research Council of Canada and the Canada Department of Energy, Mines and Resources. This study was completed during the writer's tenure as a National Research Council of Canada Postdoctorate Fellow at Dalhousie University. ALBUQUERQUE, C. A. R. and SHAW,D. M. 1972. Thallium. In: Handbook of Geochemistry (K. H. Wedepohl (Ed.)). Ch. 81, Vol. 11-1. Springer-Verlag New York Inc. BARKER, F., WONES,D. R., and SHARP,W. M. 1972. Some aspects of the chemistry, mineralogy, petrology and genesis of the Pikes Peak batholith, Colorado. Geol. Soc. Am., Ann. Mtgs., Abstracts with Programs, p. 444. BESWICK,A. E. 1973. An experimental study of alkali metal distributions in feldspars and micas. Geochim. Cosmochim. Acta., 37, pp. 183-208. BESWICK, A. E. and EUGSTER, H. P. 1968. The distribution of potassium and rubidium between sanidine and phlogopite. Geol. Soc. Am., Ann. Mtgs., Abstracts with Programs, pp. 25-26. BUDDINGTON, A. F. 1959. Granite emplacement with special reference to North America. Bull. Geol. Soc. Am., 70, pp. 671-747. W. 1967. The origin of certain granitic rocks CHESWORTH, occurring in Glamorgan Township, southeastern Ontario. Unpubl. Ph.D. thesis, McMaster Univ., Hamilton, Ontario. 1971. Metamorphic conditions in a part of the Haliburton Highlands of Ontario. Lithos, 4, pp. 219-229. CHIANG,M. C. 1965. Element partition between hornblende and biotite in the rocks from the Loon Lake aureole, Chandos Township, Ontario. Unpubl. M.Sc. thesis, McMaster Univ., Hamilton, Ontario. C ~ o o s E. , 1934. The Loon Lake pluton, Bancroft area, Ontario, Canada. J. Geol. 42, pp. 393-399. I. F. 1971.Geology of the CURRIE, K. L. and ERMANOVICS, Loughborough Lake region, Ontario with special emphasis on the origin of the granitoid rocks. A contribution to the syenite problem. Geol. Surv. Can., Bull. 199,85 p. DENECHAUD, E. B., HELMKE, P. A., and HASKIN,L. A. 1970. Analysis for the rare-earth elements by neutron activation and Ge(Li) spectrometry. J. Radioanal. Chem., 6, pp. 97-113. DOSTAL,J. 1973. Geochemistry and petrology of the Loon Lake pluton, Ontario. Unpubl. Ph.D. thesis, McMaster Univ., Hamilton, Ontario. 1975. The origin of the garnet-cordierite-sillimanite bearing rocks from Chandos Township, Ontario. Contrib. Mineral. Petrol. 49, pp. 163-175. DUPUY,C. 1970. Contribution a 1'Btude des fractionnements geochimiques des alcalins, des alcalino-terreux et du gallium au cours des processus magmatiques. Unpubl. D.S.N. thesis, Univ. Montpellier, France. I. F. 1967. Evidence bearing on the origin of ERMANOVICS, the Perth Road pluton, southern Ontario. Unpubl. Ph.D. thesis, Queen's Univ., Kingston, Ontario. FENTON,M. D. and FAURE,G. 1%9. The strontium isotopic composition of the Laramie Range syenite, Wyoming, and its bearing on petrogenesis. Geol. Soc. Am., Ann. Mtgs., Abstracts with Programs pp. 15-16. FRATTA,M. 1974. AAS determination of ppb amounts of thallium in silicate rocks. Can. J. Spectr., 19, pp. 33-37. FRATTA, M. and SHAW,D. M. 1974. 'Residence' contamination of K , Rb, Li, and TI in diabase dikes. Can. J. Earth Sci., 11,pp. 422-429. FREY,F. A., HASKIN,M. A. POETZ,J. A., and HASKIN, L. A. 1968. Rare earth abundances in some basic rocks. J. Geophys. Res., 73, pp. 6085-6098. GREEN,T. H., BRUNFELT, A. O., and HEIER,K. S. 1972. Rare-earth element distribution and K/Rb ratios in granulites, mangerites and anorthosites, LofotenVesteraalen, Norway, Geochim. Cosmochirn. Acta, 36, pp. 241-257. HART,S. R., and ALDRICH,L. T. 1966. Fractionation of K/Rb by amphiboles-implications regarding mantle composition. Science, 155, pp. 325-327. DOSTAL: THE LOON LAKE PLUTON HEATH,S. A. and-FAIRBAIRN, H. W. 1968. Srs7/SrS6ratios in anorthosites and some associated rocks. In: Origin of Anorthosite and Related Rocks (Y. W. Isachsen (Ed.)). N.Y. State Mus. Serv. Mem. 18, pp. 99-110. HEIER,K. S. and ADAMS,J . A. S. 1964. The geochemistry of the alkali metals. Phys. Chem. Earth, 5, pp. 253-381. HERRMANN,A. G. 1970. Yttrium and lanthanides. In: Handbook of Geochemistry (K. H. Wedepohl (Ed.)). Vol. 11-2. Springer-Verlag New York Inc., pp. 57-71. JACOBY, W. R. 1971. A detailed gravity study of granites in the Bancroft area, Ontario. Unpubl. manuscript, Gravity Div., Earth Phys. Br., Dept. Energy, Mines and Resources, Ottawa, 80 p. KLEEMAN,A. W. 1965. The origin of granitic magmas. J. Geol. Soc. Austr., 12, pp. 35-52. KROGH, T . E. 1964. Strontium isotope variations and whole-rock isochron studies in the Grenville Province of Ontario. Dept. Geol. Geophys., Mass Inst. Tech., 12thAnn. Rep. for 1964, U.S. AtomicEnergy Comm., pp. 73-124. KROGH,T. E. and HURLEY, P. M. 1968. Strontium isotope variation and whole-rock isochron studies, Grenville Province of Ontario. J. Geophys. Res., 73, pp. 7107-7125. LANGE,I. M., REYNOLDS,R. G. JR., and LYONS,J. B. 1966. K/Rb ratios in coexisting K-feldspars and biotites from some New England granite and metasediments. Chem. Geol., 1 , pp. 317-322. LUMBERS,S. B. 1967. Stratigraphy, plutonism and metamorphism in the Ottawa River Remnant in the Bancroft-Madoc area of the Grenville Province of southeastern Ontario, Canada. Unpubl. Ph.D. thesis, Princeton Univ., Princeton, New Jersey. MCCAMMON, B. W. 1968. A geochemical study of some igneous rocks from the Loon Lake complex using spectrographic methods. Unpubl. B.Sc. thesis, McMaster Univ., Hamilton, Ontario. MIYASHIRO, A. 1961. Evolution of metamorphic belts. J. Petrol.,2, pp. 277-311. MORSE,S. A. 1968. Syenites. In: Carnegie Inst. Washington Yearbook for 1967, pp. 112-120. NOCKOLDS, S. R. and ALLEN,R. 1953. The geochemistry of some igneous rock series. Geochim. Cosmochim. Acta, 4, pp. 105-142. 1954. The geochemistry of some igneous rock series, part 11. Geochim. Cosmochim. Acta, 5, pp. 245-285. 1956. The geochemistry of some igneous rock series, part 111. Geochim. Cosmochim. Acta, 9, pp. 34-77. PAYNE,J . G. 1966. Geology and geochemistry of the Blue Mountain nepheline syenite body. Unpubl. Ph.D. thesis, McMaster Univ., Hamilton, Ontario. REYNOLDS, R. C., JR. WHITNEY,P. R., and ISACHSEN, Y. W. 1969. K/Rb ratios in anorthositic and associated charnockitic rocks of the Adirondaks and their pet- 1345 rogenetic implication. In Origin of Anorthosite and Related Rocks (Y. W. Isachsen (Ed.)). N.Y. State Mus. Sci. Serv. Mem. 18, pp. 267-280. SAHA,A. K. 1959. Emplacement of three granitic plutons in southeastern Ontario, Canada. Bull. Geol. Soc. Am., 70, pp. 1293-1326. SAUERBREI, A. 1966. The granitic rocks of the Frontenac Axis. Unpubl. M.Sc. thesis, Queen's Univ., Kingston, Ontario. ~ H A F I Q U L L AM., H , TUPPER,W. M., and COLE,T. J . S. 1973. Radiometric ages on pegmatites from the Bancroft area, Ontario, Canada (Abstr.). In: Symposium on the Geology of the Canadian Arctic, Geol. Assoc. Can., Program with Abstracts, Saskatoon, p. 75. SHAW,D. M. 1962. Geology of Chandos Township. Ont. Dep. Mines, Geol. Rep. No. l l , 2 8 p. 1972. Origin of the Apsley gneiss. Can. J . Earth Sci., 9, pp. 18-35. SILVER,L . T. and LUMBERS, S. B. 1965. Geochronologic studies in the Bancroft-Madoc area of the Grenville Province, Ontario, Canada. Geol. Soc. Am., Ann. Mtgs., Abstracts with Programs p. 156. THORNTON, C. P. and TUTTLE,0. F. 1960. Chemistry of igneous rocks. I. Differentiation index. Am. J. Sci., 258, pp. 664-684. TUTTLE,0.F. and BOWEN,N. L. 1958. Origin ofgranite in light of experimental studies. Geol. Soc. Am., Mem. 74, 153 p. UPTON,B. G. J. 1960. The alkaline igneous complex of Kungnat field, south Greenland. Medd. Granland, 123, pp. 5-145. WHITE, A. J. R. 1966. Genesis of migmatites from the Palmer region of South Australia. Chem. Geol., 1 , pp. 165-200. WHITNEY,P. R. 1969. Variations of the K/Rb ratio in migmatitic paragneisses of the Northwest Adirondacks. Geochim. Cosmochim. Acta., 33, pp. 1203-1211. WINKLER,H. G. F. 1967. Petrogenesis of metamorphic rocks. 2nd edition, Springer-Verlag, New YorkHeidelberg-Berlin, 237 p. WINKLER,H . G. F. and VON PLATEN,H. 1961. Experimentelle Gesteins-metamorphose, V. Experimentelle anatektischer Schmelzen und ihre petrogenetische Bedeutung. Geochim. Cosmochim. Acta., 24, pp. 250-259. WYNNE-EDWARDS, H. R. 1957. The structureofthe Westport concordant pluton in Grenville, Ontaiio. J. Geol., 65, pp. 639-649. 1967. Westport map-area, Ontario, with special emphasis on the Precambrian rocks. Geol. Surv. Can., Mem. 346, 142 p. 1972. The Grenville Province. In: Variations in tectonic styles in Canada. Geol. Assoc. Can., Spec. Paper 1 1 (R. A. Price and R. J. W. Douglas (Eds .)), pp. 264-334.