Newly recognized blueschist-facies metamorphism (glaucophane-omphacite-garnet), Belvidere Mountain Complex, northern Appalachians

Preservation of blueschist- and eclogite-facies metamorphism is uncommon in the Appalachian orogen, with most known localities associated with subduction along the Laurentian margin during the early Paleozoic Taconic orogenic cycle (Fig. 1A). In New England, Cambrian to Early Ordovician Taconic subduction formed metasedimentaryrich lithotectonic terranes with numerous slivers of serpentinized ultramafic rocks and occasional blocks of barroisite greenstone, blueschist, and eclogite (Fig. 1B; Laird and Albee, 1981; Rowley and Kidd, 1981; Laird et al., 1984; Stanley et al., 1984; Stanley and Ratcliffe, 1985; Karabinos et al., 1998; Thompson and Thompson, 2003; Ratcliffe et al., 2011; Jacobi and Mitchell, 2018). Regional spatial zonation in Taconic metamorphic grade from sub–greenschist facies in northwestern Vermont to eclogite facies and ultrahigh pressure farther east reflects an eastdirected (presentday coordinates) Taconic subduction vector (Fig. 1B; Laird and Albee, 1981; Laird et al., 1984; Gonzalez et al., 2020; Honsberger et al., 2020).

The Belvidere Mountain Complex is an allochthonous ultramafic-mafic-pelitic-psammitic rock sequence in northern Vermont that occurs at the same structural level as blueschist and coesitebearing eclogite ~2 km to the north (Tillotson Peak Complex; Fig. 1C; Gale, 1986, 2007; Laird et al., 1993; Ratcliffe et al., 2011). The Tillotson Peak Complex is the only other known locality in the Appalachian orogen to preserve the assemblage glaucophane-omphacite-garnet (Laird and Albee, 1981). The Belvidere Mountain Complex and Tillotson Peak Complex are both interpreted to have been tectonized along the Taconic subduction zone (Doolan et al., 1982; Laird et al., 1984, 1993; Gale, 1986; Thompson and Thompson, 2003; Honsberger et al., 2020). Ultramafic rocks of the Belvidere Mountain Complex are well studied because they host economically important alteration mineral assemblages including talc and asbestos, the latter of which was mined at the Belvidere Mountain Complex from the late 1800s to the early 1990s (e.g., Cady et al., 1963; Chidester et al., 1978; Labotka and Albee, 1979; Gale, 1986; Carlsen et al., 2015). It is now known that Taconic subduction-related serpentinization of the Belvidere Mountain Complex ultramafic rocks was important to deep carbon cycling through the production of abiotic methane (Boutier et al., 2021). The mafic rocks of the Belvidere Mountain Complex are also of petrologic and tectonic significance because they contain the sodium-calcium amphibole, barroisite, indicating relatively high-pressure subduction zone metamorphism; however, evidence for the conditions of peak Taconic metamorphism is cryptic because of pervasive greenschistfacies retrograde overprint (Laird et al., 1984, 1993).

The purpose of this research was to carry out careful and detailed mineralogic analyses of amphibolite from the Belvidere Mountain Complex to constrain the enigmatic high-pressure conditions of Taconic subduction zone metamorphism. The preservation of a blueschist-facies mineral assemblage as inclusions in a relict garnet from the Belvidere Mountain Complex implies that evidence for highpressure, and potentially ultrahighpressure, subduction zone metamorphism may be more extensive in the Appalachians than realized at present; however, such mineral assemblages are only found through systematic analysis of retrogressed rocks (e.g., Abbott and Greenwood, 2001; Chu et al., 2016; Gonzalez et al., 2020).

Exhumed rocks of the Taconic subduction zone in northwestern New England (Fig. 1), including the Belvidere Mountain Complex, occur south along strike of Cambrian to Ordovician suprasubduction zone ophiolites in southern Quebec (Doolan et al., 1982; Tremblay and Pinet, 2016). The Belvidere Mountain Complex is a thrust slice between underlying muscovitealbite schists of the Laurentian margin (Hazens Notch and Fayston formations) and overlying graphitic phyllites (± quartz schists), chlorite–mica schists, and ultramafic rocks of the Ottauquechee, Jay, and Stowe formations, which are interpreted to represent the Taconic accretionary prism and exhumed subduction zone channel (Figs. 1B–1C; Gale, 1986, 2007; Stanley et al., 1984; Stanley and Ratcliffe, 1985; Thompson and Thompson, 2003; Ratcliffe et al., 2011; Honsberger et al., 2020). The shallowly to moderately eastdipping, westdirected Prospect Rock thrust fault accommodated Taconic imbrication of the Belvidere Mountain Complex with muscovitealbite schists and overlying phyllites, schists, and ultramafic rocks (Fig. 1B; Thompson and Thompson, 2003). A steeply dipping, longlived Paleozoic fault zone, the Burgess Branch fault, truncates the eastern margin of the Belvidere Mountain Complex and can be traced north to a zone of normal faulting in southern Quebec (St. Joseph fault; Fig. 1B; Kim et al., 1999).

The Belvidere Mountain Complex is composed of, from structural bottom to top, muscovitealbite (± garnet) gneiss, finegrained albite greenstone, medium- to coarsegrained barroisite amphibolite (± garnet) with lenses of muscovite ± garnet schist and tectonic mélange, and serpentinized, talc-rich ultramafic rocks (Fig. 1C; Gale, 1986; Ratcliffe et al., 2011). The internal structure is characterized by shallowly to moderately east-southeast–dipping, folded faults, interpreted to be Taconic thrust faults (Fig. 1C; Gale, 1986). The geology and structure of the Belvidere Mountain Complex are similar to the ultramafic-mafic-pelitic-psammitic rock sequence of the Tillotson Peak Complex, except that a rare coesite-bearing pelitic rock and numerous glaucophane- and omphacite-bearing rocks have been recognized at Tillotson Peak (Laird and Albee, 1981; Laird et al., 1993; Gonzalez et al., 2020).

Mafic rocks of the Belvidere Mountain Complex have been interpreted to be metamorphosed oceanic lithosphere that was subducted to depth along the Taconic subduction zone (Laird et al., 1984, 1993, 2001; Shaw and Wasserburg, 1984; Honsberger et al., 2019). Barroisite from garnet-barroisite amphibolite from the summit of Belvidere Mountain, at approximately the same location as the samples collected in this study (B-1 and B-2), previously yielded Cambrian to Early Ordovician ⁴⁰Ar/³⁹Ar ages of 505 ± 2 Ma (Laird et al., 1993), 490 ± 8 Ma (Laird et al., 1984), and 486 ± 31 Ma (Castonguay et al., 2012). The most precise age of 505 ± 2 Ma (Laird et al., 1993) is consistent with subduction zone metamorphism during the earliest phase of the Taconic orogenic cycle (van Staal and Barr, 2012). Muscovite from the Belvidere Mountain Complex muscovite schist unit yielded an Early Ordovician (475 ± 3 Ma) Taconic ⁴⁰Ar/³⁹Ar age (Castonguay et al., 2012). Peak metamorphic temperatures for the dated garnet-barroisite amphibolite were estimated by garnet-amphibole thermometry to be 550–650 °C (Laird et al., 1993). Calculation of one pseudosection for the sample of Laird et al. (1993) yielded 510–520 °C and 1.2–1.3 GPa for metamorphic growth of barroisite; however, those calculations did not involve ferric iron, and the solution models utilized in that study have since been superseded (Honsberger, 2015).

Petrography, electron microprobe analyses, pseudosection modeling, and mineral composition isopleth calculations were carried out on one medium-grained garnet-barroisite amphibolite sample (B-1) and one medium-grained barroisite amphibolite sample without garnet (B-2). Both samples were collected near the summit of Belvidere Mountain, close to a contact between the medium- to coarse-grained garnet amphibolite and fine- to medium-grained amphibolite ± garnet map units (Fig. 1C).

Sample B-1 is a rutile-quartz-plagioclase-titanite-garnet-chlorite-amphibole-clinozoisite schist containing relict rutile grains rimmed by titanite, chlorite pseudomorphs after garnet, chlorite inclusions in pseudomorphed garnet, and relict garnet cores with glaucophane, magnesio-hornblende, clinozoisite, and omphacite inclusions (Figs. 2A–2C). The main metamorphic fabric is dominated by preferentially oriented, retrograde chlorite that overprints an earlier, now-obscured, garnet-bearing fabric (Fig. 2A). Therefore, focus was placed on the inclusions in the relict garnet core (Fig. 2B) to investigate the metamorphic conditions for the earliest (i.e., oldest) “peak” equilibrium assemblage preserved. The “peak” equilibrium assemblage for B-1 is interpreted as clinozoisite–glaucophane–magnesio-hornblende–omphacite–chlorite–garnet–rutile–quartz. The assemblage is referred to as “peak” (with quotes) because it is possible that higher pressure and/or temperature phases that are now absent in the rock (e.g., coesite) may have been present prior to retrogression.

Sample B-2 is composed of quartz-plagioclase-rutile-titanite-chlorite-clinozoisite-amphibole, with rutile overgrown by titanite, barroisite inclusions in clinozoisite, chlorite inclusions in barroisite, and compositional zoning of fabric-forming, matrix amphibole from blue barroisite and magnesio-hornblende cores to lightblue actinolite rims (Figs. 2D–2F). Such mineral textures and inclusions in B-2 are consistent with an early “peak” equilibrium assemblage consisting of clinozoisite-barroisite-chlorite-rutile-quartz.

Representative mineral compositions and wholerock geochemical analyses are presented in Tables 1 and 2, respectively. Supplemental Text File 1¹ contains mineral nomenclature references and analytical methods, whereas Table S1 provides all electron microprobe analyses. The center of the relict garnet core in sample B-1 gave a composition of almandine (X_Fe2+ = 0.5) with grossular (X_Ca = 0.3) and minor spessartine (X_Mn = 0.1) components, whereas the outer part of the core has a slightly higher almandine component (X_Fe2+ = 0.6) and contains a minor pyrope component (X_Mg = 0.1; Figs. 2A–2B; Table 1). A clinopyroxene inclusion in the relict garnet core gave a composition of omphacite (Ca_0.5Na_0.5)(Mg_0.4Fe²⁺_0.1Fe³⁺_0.1Al_0.4)Si₂O₆, and an epidote inclusion in the same grain is clinozoisite (Figs. 2A–2B; Table 1). Glaucophane (^M4Na = 1.5 apfu; Ca/Ca + Na = 0.2) and magnesio-hornblende (^M4Na = 0.4 apfu; Ca/Ca + Na = 0.8) inclusions occur in the relict garnet core with omphacite and clinozoisite [Ca_2.1(Fe³⁺_0.4Fe²⁺_0.1^VIAl_2.2)Si_3.2O₁₂(OH); Figs. 2A–2B; Table 1]. In sample B-2, barroisite (^M4Na = 0.6 apfu; Ca/Ca + Na = 0.7) and magnesio-hornblende (^M4Na = 0.5 apfu; Ca/Ca + Na = 0.8) comprise the cores of the matrix amphibole grains, with barroisite (^M4Na = 0.8 apfu; Ca/Ca + Na = 0.6) also occurring as inclusions in matrix clinozoisite (Figs. 2C–2D; Tables 1 and S1, see ^{footnote 1}). Chlorite compositions for both samples are intermediate between clinochlore and chamosite (Mg_2.4Fe_2.4^VIAl_1.2)(Si_2.8Al_1.2)O₁₀(OH)₂.

Pressuretemperature (P-T) pseudosection and mineral composition isopleth calculations (Fig. 3) were undertaken in Perple_X (Connolly, 2005) using the analyzed wholerock compositions of the samples (Table 2; see Text File S2). The wholerock compositions of the two samples are similar, but B-1 contains slightly less SiO₂ and Fe₂O₃ and slightly more Al₂O₃, MgO, and CaO compared to B-2 (Table 2). The relative concentrations of MnO, K₂O, and Cr₂O₃ are each less than 0.3 wt% in both samples (Table 2). The analyzed wholerock compositions were assumed to have been the initial compositions prior to fractionation associated with mineral growth, which is valid for progressive metamorphism involving limited diffusion (e.g., Gaidies et al., 2006). This assumption is acceptable for interpreting P-T conditions for grain cores; however, applicability to matrix grains and grain rims is less robust because closedsystem fractionation would have continuously changed the effective wholerock composition of the sample until recrystallization was complete (e.g., Gaidies et al., 2006).

For the garnet-barroisite amphibolite sample B-1, calculations were performed in the system Na₂O-CaO-FeO-MgO-MnO-Al₂O₃-SiO₂-H₂O-TiO₂-O₂ (NCFMMASHTO) between 1.0 and 2.5 GPa and 300 and 700 °C. The “peak” equilibrium assemblage clinozoisite–glaucophane–magnesio-hornblende–omphacite–chlorite–garnet–rutile–quartz is stable at ~1.65–2.2 GPa and ~450–480 °C (Fig. 3A), with the absence of lawsonite suggesting metamorphism below ~2.0 GPa (Liou and Zhang, 2003). Based on a density of 3.0 g/cm³ for mafic oceanic crust under lithostatic pressure (Hacker et al., 2003), such a range in “peak” metamorphic pressure corresponds to a depth range of ~58–70 km. If titanite was part of the “peak” equilibrium assemblage, as opposed to rutile, the range in P-T conditions would be ~1.0–1.65 GPa and ~350–450 °C (Fig. 3A). However, this scenario seems unlikely considering that relict rutile cores are overgrown by titanite in both rocks, and there is no textural evidence for early titanite growth with garnet (Figs. 2A–2D). The compositional isopleths calculated for glaucophane, omphacite, and garnet also agree with the analyzed mineral compositions within the interpreted “peak” equilibrium assemblage field, further supporting the presence of rutile in the “peak” equilibrium assemblage (Fig. 3A; Table 1).

For the garnet-absent amphibolite sample B-2, the NCFMASHTO (MnO-absent) system was modeled between 0.5 and 2.0 GPa and 300 and 700 °C. MnO was excluded from the calculations given the negligible amounts partitioned between amphibole and chlorite in the samples. The stability of garnet is overpredicted across the P-T space for B-2 when MnO is included because the amphibole solution model does not allow for the partitioning of MnO into amphibole. The “peak” equilibrium assemblage clinozoisite-barroisite-chlorite-rutile-quartz spans a stability field of ~1.0–1.4 GPa and ~515–550 °C (Fig. 3B), which corresponds to a depth range of ~34–48 km. The calculated compositional isopleths for barroisite and clinozoisite are consistent with the analyzed mineral compositions across the “peak” equilibrium assemblage field, supporting the interpretation of the “peak” P-T conditions for B-2 (Fig. 3B; Table 1).

The P-T conditions for the relict equilibrium assemblage clinozoisite–glaucophane–magnesio-hornblende–omphacite–chlorite–garnet–rutile–quartz in sample B-1 (Fig. 3A) are consistent with blueschist-facies metamorphism (Liou and Zhang, 2003), but they may not represent the true peak P-T conditions if peak phases (e.g., coesite) were originally entrained as inclusions in garnet rims that were subsequently retrogressed (Figs. 2A–2C). The estimated blueschist-facies conditions are comparable to the P-T conditions of prograde growth of coesite-bearing garnet from the Tillotson Peak Complex (1.5–1.95 GPa, 470–560 °C; Gonzalez et al., 2020). Furthermore, minor compositional zoning in the almandine (X_Fe2+) and pyrope (X_Mg) components of the relict Belvidere Mountain Complex garnet core (Table 1) is consistent with prograde metamorphism (e.g., Gonzalez et al., 2020). Accordingly, glaucophane- and omphacite-bearing garnet of the Belvidere Mountain Complex may have grown along a prograde P-T path leading to peak Taconic metamorphism. The newly recognized blueschist-facies mineral assemblage within the Belvidere Mountain Complex begs the question as to whether ultrahigh-pressure conditions were achieved, like in the Tillotson Peak Complex, but mineralogical evidence from the Belvidere Mountain Complex is lacking at present because of intense retrogression. Regardless, blueschist-facies growth of the Belvidere Mountain Complex garnet with glaucophane and omphacite inclusions must have predated retrograde metamorphism of matrix barroisite and magnesio-hornblende, which are in textural equilibrium with chlorite pseudomorphs of garnet porphyroblasts (Figs. 2A and 2E).

The garnet-absent assemblage clinozoisite-barroisite-chlorite-rutile-quartz for sample B-2 represents high-pressure epidote-amphibolite–facies metamorphism because omphacite is absent in the equilibrium assemblage (Figs. 2D–2F; Liou and Zhang, 2003). The “peak” P-T estimate for B-2 (Fig. 3B) is nearly identical to the result calculated by Honsberger (2015) for matrix barroisite in the Belvidere Mountain Complex garnet-barroisite amphibolite sample of Laird et al. (1993). Barroisite inclusions in clinozoisite may represent retrograde, prograde, or peak metamorphism. The first two scenarios require that higher-pressure and/or higher-temperature blueschist- and/or eclogite-facies phases that are now absent in the rock (e.g., glaucophane and/or omphacite) were consumed entirely during retrogression prior to final emplacement (e.g., Palin et al., 2014). If barroisite inclusions grew along a prograde path to the blueschist-facies conditions constrained by B-1, metamorphism of B-2 would have followed a somewhat counterclockwise P-T path (Fig. 3). Alternatively, retrograde growth of the inclusions would have occurred along a clockwise P-T path from blueschist-facies conditions (Fig. 3). Regardless, the geographic proximity of samples B-1 and B-2 (Fig. 1C) suggests that the different meta-morphic conditions (blueschist and epidote-amphibolite facies) define different segments of a shared metamorphic path for the Belvidere Mountain Complex amphibolites. Sodium-rich barroisite inclusions in B-2 compared to the matrix barroisite grains (Table 1) imply retrograde growth of matrix barroisite at P-T conditions like those calculated for barroisite greenstones south along strike of the Belvidere Mountain Complex (0.65–0.85 GPa, 480–495 °C; Figs. 1B and 3B; Honsberger et al., 2017, 2020).

Garnet-barroisite amphibolite of the Belvidere Mountain Complex in the northern Appalachians preserves evidence of Taconic blueschist-facies subduction zone metamorphism, as reflected by glaucophane and omphacite inclusions in the core of a relict retrogressed garnet. Based on pseudosection calculations, blueschist-facies metamorphism occurred at ~1.65–2.0 GPa and ~450–480 °C, whereas high-pressure epidote-amphibolite–facies metamorphism occurred at ~1.0–1.4 GPa and ~515–550 °C. The blueschist-facies conditions are interpreted to reflect prograde garnet growth leading to peak Taconic metamorphism that is not quantified because of overprinting of garnet rims during retrogression. These new findings confirm a deep Taconic subduction history for the Belvidere Mountain Complex and indicate that careful observations and detailed analyses of retrogressed metamorphic rocks in the Appalachians could reveal more relict blueschist-facies mineral assemblages.

This research was inspired by field work and discussions with Jo Laird, Peter Thompson, Marjie Gale, and Wally Bothner. Many thanks go to Marc Beauchamp for carrying out electron microprobe analyses. Internal review by Tarryn Cawood and external reviews by Christopher Gerbi, Greg Walsh, an anonymous reviewer, and Associate Editor Mike Williams are appreciated and improved the manuscript. This is Natural Resources Canada contribution 20220261.