Plutons in the vicinity of the city of Saint John, as well as the Brookville Gneiss, were originally termed the "Golden Grove Intrusives" by Hayes and Howell (1937). The name was subsequently used to include plutons over a wider area in southern New Brunswick (e.g., Ruitenberg et al. 1979; Currie 1988). White et al. (1990) suggested that the term Golden Grove Intrusive Suite should be abandoned because the rocks assigned to it are not all plutonic, are not all the same age, and do not belong to a single intrusive suite. However, in recognition that a collective name is needed for plutons in the Brookville terrane, Barr et al. (2001) suggested that the name Golden Grove Plutonic Suite be redefined to refer only to the latest Neoproterozoic to Cambrian gabbroic to granitic plutons of the Brookville terrane, a recommendation followed here.

The 34 plutons that comprise the redefined Golden Grove Plutonic Suite are listed in Table 1. They include most plutons known in the Brookville terrane; exceptions are orthogneissic components of the Brookville Gneiss, the Gaytons Granite, and subsurface gabbroic, anorthositic, and granitic rocks of the Lower Coverdale Gabbroic Complex. These units are excluded because

U-Pb data indicate an older, ca. 605 Ma, age for the Brookville orthogneiss, as noted above, and a Devonian age for the Gaytons Granite and the Lower Coverdale complex (Barr et al. 2002a). Although not all plutons of the re-defined Golden Grove Plutonic Suite have been dated, a latest Neoproterozoic to Cambrian age is assumed on the basis of petrological similarities to dated plutons. Some pluton names have been changed from those used by White (1996) and White and Barr (1996) because of conflicts with preexisting unit names (L.R. Fyffe, personal communication, 2000), or because additional work clarified their inclusion in other units (Barr et al. 2001). Some plutons have been added as a result of additional work in the Pocologan and Lutes Mountain area, as described below. Because a variety of different names have been used in the past for some of these plutons, White (1996) provided, for clarification, a compilation of the previous terminology relative to current usage. Detailed maps that include the plutons in the Saint John and Pocologan areas have been published (Barr and White 2001); simplified maps are shown in Figures 2 and 3.

The plutons of the Golden Grove suite can be broadly grouped into four compositional types depending on the dominant rock type, abundance of mafic minerals, and texture: (1) gabbro, (2) diorite to granodiorite, (3) granodiorite to monzogranite, and (4) syenogranite to monzogranite. As summarized in Table 1, relative ages among some plutons of the suite can be established based on cross-cutting relationships and/or the presence of xenoliths. These observations do not suggest a clear sequence of intrusive relationships from mafic to felsic. The lack of a compositional pattern is supported also by the U-Pb ages (see below) in those cases where the ages are precise enough to indicate an order of intrusion. However, the close spatial association of the plutons suggests that they were generated in the same place at more or less the same time by more or less the same processes, but that magma evolution was parallel, as well as sequential. Such complex "compositionally expanded series" are typical of Andean-type subduction zones (e.g., Pitcher 1994).

A total of 163 chemical analyses for major and selected trace elements are available from the plutons of the Golden Grove Plutonic Suite. They include data from White (1996) and Eby and Currie (1996), as well as new data obtained in the present study (Table 2). Rare-earth element data are available for 48 samples, 14 from the present study (Table 3) and 32 from Eby and Currie (1996) and Currie (1996; written communication). The size of the data set precludes detailed discussion of chemical variations within individual plutons, and instead an overview of the chemical variations among plutons is emphasized. Because of the range in loss-on-ignition values in the samples, the major element oxide data have been recalculated to total 100% volatile-free before being plotted on the various diagrams.

The varied compositions of the Duck Lake and Indiantown gabbroic plutons are reflected in the chemical data. SiO₂ content ranges from less than 40% in ultramafic samples to 48% in gabbroic and anorthositic samples (Fig. 4). Ranges in Al₂O₃, CaO, and Na₂O correlate with the varying abundance of plagioclase; anorthositic samples have more than 20% Al₂O₃, 14—18% CaO, and 1—2% Na₂O (Figs. 4b, e, f). Ultramafic (lowest SiO₂) samples have higher Fe₂O₃ and MgO (Figs. 4c, d) and low Al₂O₃, CaO, and Na₂O (Figs. 4b, e, f). Most samples are low in TiO₂, K₂O, and P₂O₅ (Figs. 4a, g, h). The low TiO₂ and P₂O₅ are in marked contrast to the Lower Coverdale gabbroic complex sampled in drill holes near Moncton (Fig. 1), in which TiO₂ and P₂O₅ contents exceed 20% and 8%, respectively, in some gabbroic samples (Barr et al. 2002d).

The gabbroic samples with highest SiO₂ contents overlap with the most mafic (lowest SiO₂) samples from the dioritic plutons. These overlapping samples have generally similar abundances of most major element oxides and in the cases of Al₂O₃, MgO, CaO, Na₂O, and K₂O the gabbroic samples lie on reasonably linear trends with the dioritic samples (Figs. 4b, d, e, f, g), suggesting a genetic relationship. Trends are less continuous in the cases of TiO₂, Fe₂O₃ ^t, and P₂O₅ (Figs. 4a, c, and h), but those more erratic variations could result from fractionation of titaniferous magnet-ite, ilmenite, and apatite.

Trace elements such as Rb and Zr that are incompatible in mafic minerals and plagioclase are generally low in the gabbroic samples (Figs. 5a, b). Conversely, compatible elements show wide variation (e.g., V, 18 to 437 ppm; Cr, 36 to 2926 ppm, and Ni, 45 to 838 ppm) and positive correlation with TiO₂ (Fig. 5f; one sample with 437 ppm V and 1.36% TiO₂ is off the scale of the figure) and Fe₂O₃t and MgO (not shown).

REE concentrations are low and, relative to chondritic values, show slight light REE enrichment (Fig. 6a). All four samples show slight to moderate enrichment in Eu, suggesting accumulation of plagioclase in the analyzed samples.

The large range in chemical compositions in the dioritic to granodioritic plutons corresponds well with the range in modal mineralogy. SiO₂ contents vary from less than 50% in dioritic samples to over 75% in the most granitic parts of some plutons, but the majority of the samples are intermediate in composition, with between 60% and 70% SiO₂ (Fig. 7). Most major element oxides show a strong negative correlation with SiO₂ (Figs. 7a, b, c, d, e, h), with the exceptions of Na₂O, which is relatively constant at 3—4% (Fig. 7f), and K₂O, which shows a positive correlation with SiO₂ (Fig. 7g). Trace elements such as Ba, Rb, and Sr that are compatible in feldspars, generally the most abundant minerals in these samples, show trends similar to the major oxides, as illustrated by Rb which shows strong correlation with SiO₂ and K₂O (Figs. 5a, c). The major and trace element trends are consistent with fractional crystallization of mafic minerals and feldspars as the major cause of chemical variation within this suite of plutons. In contrast, incompatible elements such as Zr show little co-variation with SiO₂ or K₂O (Figs. 5b, d), but positive correlation with some other elements such as Y (Fig. 5e). Positive correlation between V and TiO₂ follows a different trend than that in the Duck Lake and Indiantown gabbroic samples (Fig. 5f). This difference suggests that fractionation in the dioritic to granodioritic suite involved minerals with lower V relative to Ti than the gabbroic suite, or that the parental magmas for the gabbroic suite contained higher V relative to Ti.

Three samples from mafic enclaves in these plutons generally show chemical similarities to the dioritic samples with lowest SiO₂ contents (Figs. 5, 7). This similarity supports the possibility that the enclaves are cognate, and represent concentrations of minerals crystallized from the same magmas as their host rocks. This interpretation is further supported by REE data from one of the enclaves, which displays a chondrite-normalized pattern very similar to that of the dioritic rocks, but slightly less evolved (Fig. 6a). The REE patterns for the dioritic and granodioritic suite as a whole show light REE enrichment (La between 45 and 150 times chondritic values) and nearly flat heavy REE at 6.5 to 25 times chondritic values). The patterns are nearly parallel to one another and show slight to moderate negative Eu anomalies (Fig. 6b). These patterns are consistent with fractionation of plagioclase, amphibole, and biotite, as suggested by major and trace element variations.

Plutons in this group are dominated by granodiorite and monzogranite, consistent with their SiO₂ contents which are mainly between 67% and 71%, with rare more syenogranitic samples ranging up to 76% SiO₂ (Fig. 7). One sample from the Hanson Stream Granodiorite analyzed by Eby and Currie (1996) contains about 60% SiO₂ and may be from a mafic enclave. The samples show compositional overlap with data from the more abundant dioritic to granodioritic plutons and together the two groups of plutons form strong trends, with the exception of the samples from the Fairville and Chalet Lake plutons. The latter samples show trends of higher TiO₂, Fe₂O₃ ^t, and K₂O and lower Al₂O₃, MgO, and CaO compared to the other plutons (Fig. 7). Trace elements show similar differences; for example, many of the analyzed samples from the Fairville and Chalet Lake plutons have higher Rb, Zr, and Y (Figs. 5a, b, d, e) and lower V (Fig. 5f). Total REE concentrations are higher, especially the heavy REE (Fig. 6c). In contrast, the other granodioritic and monzogranitic plutons have REE patterns similar to those of the dioritic to granodioritic samples.

Analyzed samples from the syenogranitic to monzogranitic plutons have more than 70% SiO₂, with the exception of 2 samples from the more mafic parts of the composite Musquash Harbour Pluton (Fig. 7). They generally have higher SiO₂ contents than samples from the granodioritic and monzogranitic plutons, although there is considerable overlap. However, on many of the major element variation diagrams, the syenogranitic samples form linear trends with samples from the Fairville and Chalet Lake plutons, trends that involve higher TiO₂, Fe₂O₃t, and K₂O and lower Al₂O₃, MgO, and CaO compared to the trends of the other plutons. Trace elements show similar variations; the syenogranitic and monzogranitic suite of samples has higher Rb, Zr, and Y (Figs. 5a, b, e) and lower V (Fig. 5f). They tend to form scattered trends with the samples from the Fairville and Chalet Lake plutons that differ from those in the remaining samples, such as a negative correlation between Zr and SiO₂ (Fig. 5b). The REE values are higher and like those in the Fairville Granite, although they display wider variation in heavy REE and more pronounced negative Eu anomalies than the Fairville Granite samples (Fig. 6d). Such differences suggest that these plutons may be related by fractionation of K-feldspar and zircon.

In contrast to the differences displayed by the syenogranitic to granodioritic samples from the Musquash Harbour Pluton, two quartz diorite samples and a granodiorite sample from that pluton are chemically like dioritic samples from the dioritic to granodioritic suite, and generally plot within the trends defined by those samples (Figs. 5, 7). The similarity is also shown by the REE data from one of these samples, which is identical to the REE pattern of the other dioritic samples (Fig. 6a). These data suggest that the dioritic parts of the Musquash Harbour Pluton may not be cogenetic with its more voluminous granitic parts, and may instead be related to the dioritic — granodioritic plutons of the Brookville terrane.

A field for 5 analyzed samples from the Dipper Harbour volcanic unit is shown on the various chemical plots for comparison with the granitoid rocks. The samples show limited chemical variation, and are similar to samples from the syenogranitic plutons, with about 77—78% SiO₂ (Fig. 7). Like those plutons they show elevated contents of Rb, Zr, and Y, and low V and TiO₂ (Fig. 5), and similarly have within-plate and A-type characteristics (Fig. 8). The REE pattern of one sample (NB92-9073; Table 3) is similar to those of the syenogranitic samples (Fig. 6d). These chemical similarities support a comagmatic relationship between the syenogranitic plutons and the volcanic rocks, also suggested by their spatial association and U-Pb ages, as noted above.

Most of the analyzed samples plot on a clear calc-alkaline trend on an AFM diagram (Fig. 8a), although some samples from the Fairville, Chalet Lake, and syenogranitic plutons plot away from the trend because of their higher Fe contents, and most gabbroic samples plot toward the MgO corner. Most of the dioritic to granodioritic samples, as well as the granodioritic and monzogranite samples, plot in the volcanic arc field on Zr-TiO₂ and Rb-Y+Nb tectonic setting discrimination diagrams (Figs. 8b, c). However, on both of these diagrams, the syenogranitic and monzogranitic samples, and those from the Dipper Harbour volcanic unit and the Fairville and Chalet Lake plutons, are offset toward the within-plate fields. Samples for which Th, Yb, and Ta data are available plot mainly in the active continental margin field (Fig. 8d). Most samples have "I-type" affinities, although the syenogranitic and monzogranitic samples, as well as those from the Dipper Harbour volcanic unit and Fairville and Chalet Lake plutons, mainly plot in the A-type fields (Figs. 8e, f), consistent with their "within-plate" tendencies.

The five U-Pb dates reported here were done by three of the authors in different laboratories, although the methodology was similar. The Fairville Granite and Ludgate Lake Granodiorite samples were dated by C. White in the laboratory of G. Dunning at Memorial University of Newfoundland, by methods described in White (1996). The Lutes Mountain Diorite and McCarthy Point Granodiorite were dated by B. Miller at the University of North Carolina, Chapel Hill, using methods similar to those described by Ratajeski et al. (2001). The sample from the Duck Lake Pluton was dated by M. Hamilton at the Geological Survey of Canada, Ottawa, using methods described by Barr et al. (2000). The analytical data for all five samples are presented in Table 4, together with UTM co-ordinates of sample locations.

A pegmatoid area within texturally and compositionally varied gabbro near the northwestern margin of the Duck Lake Pluton was sampled for dating. Dated sample DL97-1 was mostly coarse-grained plagioclase-rich gabbro, locally with abundant hornblende and minor quartz. Zircon grains recovered from the sample were mostly colourless to very pale brown and subhedral to anhedral in shape, the latter suggesting relatively late magmatic growth. Individual grains ranged from approximately 75—150 µm in maximum dimension. All grains were relatively clear and inclusion-free. However, many zircon grains showed minor, thin overgrowths of probable metamorphic origin. All grains were therefore given extensive air abrasion treatment (up to ca. 35 hours) to remove all optical signs of rims, and the best quality grains were subsequently re-picked from these populations, for chemistry.

Analysis of four fractions, each comprising between 52—70 grains, yielded results that are tightly clustered, lying on or immediately below concordia, and have 207Pb/206Pb ages which range narrowly between 538.9 and 540.4 Ma (Table 4, Fig. 9a). The minor discordance displayed by the data (0.3—0.9%) can be modelled by recent (modern-day) Pb-loss. The zircon data collectively define a weighted mean 207Pb/206Pb age of 539.6 ± 1.2 Ma, which we interpret to represent the igneous crystallization age of the Duck Lake Gabbro.

Sample NB92-9012 was collected from the Fairville Granite in a road cut south of Green Head Island. It consisted of coarse-grained inequigranular biotite monzogranite, typical of the pluton. It yielded a zircon population composed of colourless to very pale yellow euhedral dipyramidal prisms, with length/ breadth ratio of about 3.3. The grains showed good to excellent clarity, with clear tube- or bubble-like inclusions and no visible evidence of inherited cores. Two abraded zircon fractions (Z1, Z3) were hand picked, avoiding any grains with inclusions, and a third abraded fraction (Z2) contained minor inclusions. Analyses of Z1 and Z3 are slightly discordant (<3.3%) with 207Pb/206Pb ages of ca. 570 Ma and 560 Ma, respectively (Table 4). Analysis of fraction Z2 is 11.4% discordant and has a significantly older 207Pb/206Pb age of ca. 631 Ma (Table 4). The three fractions define a simple discordia line with lower and upper intercept ages of 548 ± 2 Ma and 1997 +280/-215 Ma, respectively (Fig. 9b). The lower intercept age is the best estimate of the minimum age of emplacement of the Fairville Granite. The upper intercept indicates the presence of a significant component of inherited zircon with an average Early Proterozoic age.

Dated sample NB00-11 was collected from a large quarry at the summit of Lutes Mountain north of Moncton. It was a medium-grained diorite that consisted mainly of plagioclase and blue-green amphibole, with minor interstitial quartz and accessory titanite, zircon, apatite, and opaque minerals. Five fractions of zircon were analyzed from the sample. Each of three fractions consisted of three or four grains of medium-sized (100 x 40 µm) prismatic zircons. The other two fractions (35 and 110 µm) consisted of multi-faceted equant grains (Table 4). One fraction is concordant at 542 Ma and two others are nearly concordant. Two additional analyses are discordant along a recent Pb-loss line. One fraction has large U-Pb errors because of an imprecise uranium analysis. All five analyses fit on a discordia line with an upper intercept of 542.2 +1.8/-1.4 Ma and a lower intercept suggestive of recent Pb-loss. We interpret the upper intercept age to be the time of crystallization of the Lutes Mountain Diorite.

Sample NB91-9010 was collected from the Ludgate Lake Granodiorite at a roadcut on highway 1, about 300 m west of Ludgate Lake. It was a medium-grained inequigranular biotitehornblende granodiorite, and contained two morphologically distinct zircon populations. The most abundant grains (>60%) are colourless, euhedral, needle-shaped, dipyramidal simple prisms, with an average length to breadth ratio of about 6. They exhibited excellent clarity with minor clear tubes and bubbles as inclusions, and no visible cores. The other 40% of the zircon grains are stubby to slightly elongate, euhedral, clear, colourless multifaceted dipyramids with an average length to breadth ratio of 2. They have clear tubes and bubbles as inclusions, and no visible cores. Three fractions were analyzed, one from the needle-shaped prisms (Z1) and two from the stubby population (Z2 and Z3). The analyses are clustered and slightly discordant (<2%) with 207Pb/206Pb ages of ca. 548 to 544 Ma.

The sample also contained titanite, light amber to dark brown, clear to slightly cloudy with an anhedral to subhedral shape, and no visible inclusions or cores. Two fractions (T1 and T2) were analyzed, of which T1 was more abraded than T2. Both fractions are slightly discordant, although T1 yields a 207Pb/206Pb age of ca. 545 Ma. The zircon and titanite fractions together define a discordia line with an upper intercept of 546 ± 2 Ma, which is interpreted to be the crystallization age of the Ludgate Lake Granodiorite. The lower intercept of ca. 30 Ma is uncertain due to the length of projection, but probably reflects recent Pb loss. The agreement of the titanite age and the upper intercept age suggests rapid cooling, at least through the closure temperature of titanite, as also confirmed by 40Ar/39Ar ages of hornblende (see below).

Dated sample NB99-5 was collected at the shoreline on the western side of McCarthy Point in Pocologan Harbour. It was a medium- to coarse-grained biotite-hornblende granodiorite typical of the pluton, and contained zircon grains of several different morphologies. In general, smaller grains were clear and free of inclusions, whereas larger grains tended to be slightly metamict. Seven zircon fractions were analyzed, of which five consisted of one or two grains and the remaining two were multi-grain fractions. None of the fractions shows evidence of inheritance. The metamict grains are more discordant than the clear grains, although both types were highly abraded. The seven fractions form a discordant trend with an upper intercept of 528 +4/-3 Ma, which we interpret to represent the time of crystallization of the pluton.

Including the data presented here and previously published dates, ten U-Pb (zircon) ages have been obtained from the Golden Grove Plutonic Suite, although the age from the Musquash Harbour syenogranite has a large error associated with it (Table 1; Fig. 10). The spread of ages between about 550 Ma and 525 Ma is real, in that the error ranges of the older and younger dates do not overlap (Fig. 10). No strong pattern of age in comparison to pluton composition or location is apparent. For example, although the two youngest ages are both from plutons in the southwestern part of the terrane, one of the oldest plutons (Harvey Hill) is in the same area.

Additional constraints on the minimum ages of pluton emplacement are provided by 40Ar/39Ar ages from hornblende in the plutons, in some cases by more than one age determination from the same pluton. Taking errors into account, the 40Ar/39Ar cooling ages show a range very similar to that of the U-Pb crystallization ages, from close to 550 Ma to about 520 Ma. They indicate that the plutons cooled rapidly through at least the argon retention temperature in hornblende (ca. 525°C; McDougall and Harrison 1988). Hornblende and phlogopite ages from the host rocks of the pluton (both the Green Head Group and the Brookville Gneiss) show similar ages (Fig. 10), consistent with the interpretation that they record pervasive contact metamorphism. The similarity in zircon and titanite U-Pb ages and 40Ar/39Ar cooling ages suggests that the plutons were emplaced are relatively shallow depth and cooled rapidly. Muscovite 40Ar/39Ar ages are somewhat younger ca. 520 Ma to 505 Ma, perhaps reflecting a decrease in the rate of cooling through the argon retention temperature for muscovite (ca. 325°C; Snee et al. 1988). Alternatively, the younger ages may reflect a subsequent reheating event.

The similarity in age at ca. 550 Ma between the Dipper Harbour volcanic unit and the syenogranitic plutons is consistent with a comagmatic relationship between them. The relationship is also suggested by the chemical similarities discussed above.