Rapid magma ascent and formation of the Águas Belas–Canindé granitic batholith, NE Brazil: evidence of epidote dissolution and thermobarometry

Mineral chemistry and intensive parameter estimates for the Major Isidoro (626 Ma) and Monteirópolis (627 Ma) magmatic epidote-bear-ing granitic plutons, emplaced along the Jacaré dos Homens transpressional shear zone (JHSZ), Borborema Province, are focused in this study. These plutons consist of medium-to-coarse grained equigranular to porphyritic tonalite to granite that show abundant dioritic enclaves. These granites contain biotite (Fe# 0.44 to 0.55), Fe-edenite (Major Isidoro), hastingsite (Monteirópolis), titanite, and epidote that often show allanite core as key mafic mineral phases. Pistacite molecular content in epidote is in the interval of 27 to 31 mol%, presenting TiO 2 < 0.30%, typical for magmatic epidote. Estimated intensive parameters reveal crystallization at 6.5 ± 1 (Major Isidoro) and 4.7 ± 0.6 kbar (Monteirópolis), temperatures from ~940°C (near-liquidus) to 675 ± 35°C (near-solidus) and oxidizing conditions. Partial corrosion of epidote took place during 15.6 to 32 (Major Isidoro) and 27–49 years (Monteirópolis), corresponding to rather high magma ascension rates of 365 to 750 and 395 to 635 m.years -1 , respectively. The JHSZ likely favored upward magma transport at the Sergipano and Pernambuco-Ala-goas domains boundary, during the onset of the Brasiliano orogeny (650–620 Ma). ~700°C. Our results point to crystallization within the range for typical high-temperature titanite (> 700°C), which is consistent with the early crystallization of euhedral titanite in these plutons.


INTRODUCTION
Mineral and whole-rock chemistries constitute an important way to assess intensive crystallization parameters and to give information on the nature, source, and evolution of the magmas. These parameters (P-T-fO 2 -time path) are of first-order importance to comprehend the physicochemical evolution of granitic magmas. Studies on mineral chemistry of natural and experimental assemblages showed that ferromagnesian and coexisting mineral phases constitute a tool for estimating intensive crystallization parameters of granites (e.g., Anderson et al. 2008). For example, biotite composition has been largely used as a redox and tectonomagmatic indicator (e.g., Wones and Eugster 1965, Nachit et al. 1985, Abdel-Rahman 1994, Shabani et al. 2003among others). Calcic amphiboles are of special interest due to their compositional diversity and common occurrence, which constitute a good potential to investigate magmatic processes (Putirka 2016, and references therein). This phase has been successfully employed as a geobarometer and, together with plagioclase (the pair amphibole-plagioclase), as a powerful geothermometer (e.g., Hammarstron and Zen 1986, Hollister et al. 1987, Schmidt 1992, Holland and Blundy 1994, Mutch et al. 2016. Titanite shows potential to quantitative estimates of pressure and temperature (e.g., Enami et al. 1993). Magmatic epidote (mEp) was initially described as indicating high-pressures of crystallization in intermediate calc-alkaline granites (> 0.5 GPa; Zen and Hammarstrom 1984). However, the crystallization of this phase is strongly dependent on oxygen fugacity and bulk magma composition. Thus, at fO 2 buffered by hematite-magnetite (HM) the stability of epidote is shifted to a lower pressure (> 0.3 GPa; Schmidt and Thompson 1996, Ferreira et al. 2003, Schmidt and Poli 2004. Furthermore, this phase has been successfully employed to estimate the ascension rate of epidote-bearing magmas (e.g., Sial et al. 1999, 2008, Brasilino et al. 2011. The Borborema Province (BP), in northeastern Brazil, comprises several intermediate plutonic rock associations that were recognized as containing epidote along with the main mafic coexisting mineral assemblage (biotite, amphibole, titanite) (e.g., Sial 1990, 1993, Brasilino et al. 1999, 2008, Long et al. 2005, 2019, Campos et al. 2016. Therefore, it constitutes an important site for mineralogical studies using mineral chemistry, which allows estimating intensive crystallization parameters, including magma ascension rate. However, there is limited work regarding intensive crystallization parameters of plutonic rocks in the BP, especially in the Pernambuco-Alagoas Domain, where the presence of mEp was recently reported (cf. Silva et al., 2015, 2016, 2019, Silva 2017).
Rapid magma ascent and formation of the Águas Belas-Canindé granitic batholith, NE Brazil: evidence of epidote dissolution and thermobarometry

GEOLOGICAL BACKGROUND
The Borborema Province, northeastern Brazil, is characterized by widespread Neoproterozoic-Cambrian granitic magmatism whose emplacement in many places is associated with shear zones (Santos and Medeiros 1999, Neves et al. 2006, Van Schmus et al. 2008, Amorim et al. 2019, Guimarães et al. 2004, Lima et al. 2017. The province is characterized by a mosaic of shear zones that are used to identify six main domains; one of them named Pernambuco-Alagoas (Fig. 1). Within this domain three granite series were identified: calc-alkaline, high-K calc-alkaline, and shoshonitic (e.g., Silva Filho et al. 2002, 2016. Silva Filho et al. (2014) divided the Pernambuco-Alagoas Domain into the Garanhuns, Água Branca, and Palmares subdomains. Migmatitic gneiss and granitic rocks of the Garanhuns sub-domain (northern portion) are derived predominantly from older (Paleoproterozoic) sources, whereas the others subdomains (southern portion) contain rocks derived from substantially younger (Tonian) sources. The Buíque-Paulo Afonso, Águas Belas-Canindé, and Ipojuca-Atalaia granitic batholiths (Silva Filho et al. 2002), each consisting of various plutons, which intruded Paleoproterozoic gneisses, represent these younger sources of the southern portion of the Água Branca and Palmares subdomains. This study focuses on part of the Águas Belas-Canindé (Água Branca subdomain).

Field relationships and petrography
The studied plutons constitute NE-SW elongate intrusions along the boundary of the Pernambuco-Alagoas and Sergipano domains, limited to the south/southeast by the Jacaré dos Homens transpressional shear zone (JHSZ) (Fig. 2) (De , Mendes et al. 2009, Lima et al. 2014. The Major Isidoro pluton is in sharp contact to the north with the Poço da Cacimba, and to the west with the Monteirópolis plutons (Silva et al. 2016, Lima 2019. The Major Isidoro pluton presents two main facies: biotite granite and epidote-amphibole-biotite tonalite, which consist of pink-gray leucocratic to mesocratic, coarse-grained to porphyritic with alkali-feldspar up to 5 cm in length. It shows a well-developed magmatic low angle dipping foliation that usually strikes northwest. However, in some samples, superposition of high-temperature   Silva Filho et al. 2002, 2016. solid-state deformation expressed by chessboard extinction in quartz, development of myrmekitic intergrowths around alkali feldspar, and static recrystallization of feldspar grains occur. Elliptical diorite enclaves up to 0.5 meters long occur everywhere into this pluton; in some places, they appear along with the granite magmatic foliation, implying that they represent disrupted syn-plutonic dikes.
The Major Isidoro granite has undergone partial melting, segregation, and ductile deformation that suggest highstrain synkinematic emplacement. This pluton was emplaced coevally with an amphibolite facies metamorphic event in the region. In some parts, the granite is migmatized, where diatexite migmatite (cf. Sawyer 2000Sawyer , 2008, which shows up to 1-meter heterogeneous schlieren features, are more common than metatexite migmatite; in this later, magmatic foliation is preserved. The schlierens in the diatexites are gray, characterized by biotite-rich domains separated by quartz-feldspar domains. The Monteirópolis pluton consists of leucocratic mediumto coarse-grained amphibole-biotite alkali feldspar granodiorite to granite; porphyritic granite that displays up to 4.5 cm long zoned alkali feldspar crystals is also found. This pluton shows magmatic foliation that has dominantly gentle dips (usually < 30°, but sometimes up to 55°) to the northwest. Diorite enclaves and amphibole-rich clots, typically a few centimeters to about 25 cm long, are present mainly in the eastern portion of the pluton. These enclaves have sharp contacts with the host granite and show regular, generally oval, but occasionally lobate, with rare diffuse contours.
The mineral assemblage of quartz + alkali feldspar + plagioclase + biotite ± hornblende + titanite ± alanite ± magnetite ± ilmenite± epidote is common within both investigated plutons. Biotite and amphibole are the main mafic phases, and together with the shape-preferred orientation of alkali feldspar define the magmatic/metamorphic foliation of these plutons. Minor saussurite is observed as products of post-magmatic transformations of feldspars. Silva et al. (2015Silva et al. ( , 2016 presented an integrated study of Sr, Nd, Pb, and O isotopes for the studied intrusions. U-Pb SHRIMP zircon crystallization ages of the Major Isidoro and Monteirópolis batholiths are respectively 626.6 ± 4.1 and 625.8 ± 3.7 Ma. Inherited zircon cores from the Major Isidoro yielded ages varying from 800 to 1,000 Ma. These crystallization ages point to an emplacement of the Major Isidoro and Monteirópolis batholiths to early stages of the Brasiliano orogeny.

ANALYTICAL PROCEDURES
All samples for whole-rock chemical analyses were crushed in a stainless steel jaw crusher and split to < 74 μm grain size in a chromium steel ring mill. The Loss On Ignition (LOI) was determined on one gram of a pre-dried sample by igniting to 1,000°C for two hours. Major and some trace elements (Sr, Zr, Nb, Y, Rb, Ba, and Ni) were determined on fused discs of sample mixed with lithium tetraborate flux (ratio sample: flux = 1:5 fused in a Pt-Au crucible) with a fully automated Rigaku RIX-3000 XRF spectrometer at the NEG-LABISE, Universidade Federal de Pernambuco, Recife, Brazil. The calibration curves were prepared by analyses of international reference materials (AC-E, AL-I, AN-G, BE-N, IF-G, and MA-N from the  International Working Group (IWG),  Mineral chemistry analyses were performed at the Electron Microprobe Laboratory, Universidade de Brasilia, Brazil, using a Superprobe JEOL JXA-8230 equipped with WDS, with an accelerating potential of 15 kV, a current of 10 ηA, and a 5μ diameter beam. WDS spot analyses of the main rock-forming minerals were performed using the following standards and X-ray lines: albite (Na Kα), forsterite (Mg Kα), topaz (F Kα), microcline (Al Kα, Si Kα, K Kα), andradite (Ca Kα), vanadinite (V Kα, Cl Kα), MnTiO 3 (Ti Kα, Mn Kα), Cr 2 O 3 (Cr Kα), SrSO 4 (Sr Lα), NiO (Ni Kα), andradite (Fe Kα), and BaSO 4 (Ba Lα). Counting times were 10s on peak and 5s on the two background positions for all elements. Matrix effects corrections were computed following the PRZ procedure and raw data reduction was done with Armstrong software from JEOL. Cationic proportions and structural formulae of plagioclase, amphibole, titanite, biotite, and epidote were calculated based on 32, 23, 5, 22 and 12.5 oxygens in the formula unit, respectively (Deer et al. 2013). Tables 1 to 5 contain representative microprobe chemical analyses.

Amphibole
Analyzed hornblende grains (Tab. 2) classified according to nomenclature scheme proposed by Leake et al. (1997Leake et al. ( , 2003 are members of the calcic group in which (Ca + Na) B ≥ 1.50 and Na B ≤ 0.50. Amphibole grains from the Major Isidoro granite and enclaves show Si content from 6.25 to 6.43 atoms per formula units (apfu), Al VI contents in the range of 0.25 to 0.534 apfu, and Ti content in the 0.04 to 0.12 apfu range. Amphiboles from granites and enclaves are, respectively, hastingsite and Mg-hastingsite (Fig. 4). Amphibole grains from the Monteirópolis pluton have lower Si contents than those in the Major Isidoro granite, in the range of 6.56 to 6.66 apfu, Al VI contents in the range of 0.17 to 0.29 apfu, and Ti contents in a narrow range from 0.06 to 0.11 apfu; they are classified as Fe-edenite to edenite (Fig. 4). Based on the new classification and nomenclature scheme for the amphiboles (Hawthorne et al. 2012), all analyzed hornblendes classify as pargasite. Amphibole compositions were plotted (Figs. 5A and 5B) in order to test the role of Edenite and Tschermak exchanges. The relatively homogeneous compositions suggest that Edenite (Al VI < 0.2 apfu) and Tschermak (Al Tot < 0.2 apfu) exchange mechanisms play a minor role in the evolution of the two plutons.

Titanite
Titanite grains from the studied plutons show very low Al + Fe 3+ (< 0.12; Tabs. 3A and 3B). The composition of titanite depends on temperature and pressure (Enami et al. 1993). These authors suggested that titanite crystals having Al + Fe 3+ < 0.35 apfu indicate crystallization at temperatures greater than ~700°C. Our results point to crystallization within the range for typical high-temperature titanite (> 700°C), which is consistent with the early crystallization of euhedral titanite in these plutons.

Biotite
Biotite from the Major Isidoro granites has higher Al tot contents (2.67 to 2.94 apfu and Fe# (Fe/(Mg + Fe) of 0.50 to 0.58) compared to biotite in Monteirópolis granites, with Al tot varying from 2.49 to 2.85 apfu, and Fe# from 0.44 to 0.55 (Tabs. 4A and 4B). These compositions lying approximately between the annite and phlogopite end-members (Fig. 6). Such compositions are in the typical range from orogenic calc-alkaline granites when compared with those presented by Nachit et al. (1985) and Abdel-Rahman (1994) (Figs. 7, 8A and 8B). Biotite crystals from both plutons have fluorine contents dominant over chlorine.

Intensive crystallization parameters
Near-solidus temperature Blundy and Holland (1990) proposed a hornblende-plagioclase thermometer with estimated uncertainties of ± 75°C, in a temperature range of 500-1,100°C, for quartz-saturated rocks. In a later work, Holland and Blundy (1994) extended the formulation of Blundy and Holland (1990) to embrace a broad range of bulk compositions (e.g., natural garnet amphibolites). The calibration based on the equilibrium: edenite + albite = richterite + anorthite (reaction B of Holland and Blundy 1994) shows lower values of temperature than those estimated using reaction A of Holland and Blundy (1994), and are considered to be the most reliable because they more precisely reproduce the T estimated by other independent methods such as garnet-hornblende and clinopyroxene-hornblende (Anderson 1996, Bachmann andDungan 2002). This thermometer works well in the ranges of 400-1,000°C (± 35-40°C) and 1 to 15 kbar over a broad range of bulk compositions. Molina et al. (2015) calibrated an empirical amphibole/liquid Mg partitioning thermometer using regression methods and observed that their T estimates are consistent with those of the edenite-albite-richterite-anorthite thermometer of Holland and Blundy (1994). Near-solidus hornblende-plagioclase temperatures (reaction B of Holland and Blundy 1994) for the Major Isidoro and Monteirópolis plutons range from 680 to 720 and from 660 to 700°C, respectively.    Watson and Harrison (1983) developed a solubility model based on the relationship between zircon crystallization and melt composition, and defined a saturation behavior of Zr as a function of temperature and magma composition. This model could be used when zircon was one of the earliest minerals to crystallize and assumes that zircon was not a cumulate phase, xenocrystic, or inherited from the source region. Sensitivity tests reinforce previous thoughts (see Watson & Harrison 1983) and indicate that temperature and composition are the two dominant controls on zircon solubility in crustal melts with no observable effects due to pressure (up to 25 kbar) or variable water content (Boehnke et al. 2013). Estimates of the near liquidus temperature can be obtained by the Equation 1, using whole-rock analyses: (1)

Near-liquidus temperature
According to this method, the calculated temperatures for the Major Isidoro pluton vary from 740 to 868°C (avg. 818°C) and to the Monteirópolis pluton temperature ranges from 763 to 856°C (avg. 805°C). Very low Zr saturation temperatures for some samples seem to underestimate the liquidus temperature with values slightly higher than those obtained to near-solidus temperature. The studied plutons present zircon cores inherited from the source, which interfere in the obtained Zr saturation temperatures (mean T Zr for inheritance-rich granites of 766°C were found by Miller et al. 2003). Green and Watson (1982) and Harrison and Watson (1984) related P 2 O 5 contents at a given silica value to temperature at which that composition may be expected to crystallize apatite. This model could be used for silica compositions between 45 and 75%, and 0 and 10% water, and for the range of pressures expected in the crust (up to 25 kbar; Green and Adam 2002). Iron content or oxidation state of the liquid play insignificant effects on phosphate saturation (Tollari et al. 2006). According to Harrison and Watson (1984), the following expression can be used to estimate the minimum liquidus temperature: T (°C) = {[8,400 + 26,400(SiO 2 -0.5)]/[ln(42/P 2 O 5 ) + 3.1 + 12.4(SiO 2 -0.5)] -273.15}. Using this approach, the granites of the Major Isidoro pluton have near-liquidus temperatures that vary between 863 and 1,000°C (average = 936°C) while granites from the Monteirópolis pluton present a wider temperature range, from 814 to 1,023°C (Tab. 6). Apatite is an early phase relative to zircon in the crystallization sequence from both plutons, and apatite exhibits significantly higher saturation temperatures (e.g., Anderson et al. 2008, Naranjo andVlach 2018) with an acceptable near-liquidus to solidus temperature span of ~940 to ~700°C.

Amphibole barometry
Several empirical and experimental igneous barometers based on Al-in-hornblende were proposed by Hammarstron andZen (1986), Hollister et al. (1987), Schmidt (1992), Anderson and Smith 1995, Anderson 1996, Holland and Blundy (1994, and Mutch et al. (2016) to granitic rocks that have the mineral assemblage plagioclase + K-feldspar + quartz + hornblende + biotite + titanite + magnetite ± ilmenite + melt + fluid phase in equilibrium. The reported revised expression for the Al-in-hornblende barometer of Anderson and Smith (1995) incorporating the effect of temperature from the hornblende-plagioclase thermometer of Holland and Blundy (1994) and using the experimental data from Johnson and Rutherford (1989) and Schmidt (1992) Accordingly, when applied to the epidote-bearing Major Isidoro and Monteirópolis plutons, pressures of 5.27-7.71 and 4.3-5.34 kbar, respectively, are obtained. These results of the Al-inhornblende barometer of Schmidt (1992) and Anderson and Smith (1995) are portrayed in Table 2. These values are generally within the uncertainty of ± 0.6 kbar. These pressures of crystallization are within the lower limit of stability of magmatic epidote (0.3 to 0.7GPa; Hammarstrom 1984, Schmidt andPoli 2004).

Oxygen fugacity
Oxygen fugacity by far exerts the strongest control on mafic silicate mineral chemistry (Anderson andSmith 1995, Anderson 1996). This parameter is difficult to estimate in silicic rocks, especially those that contain only one Fe-Ti oxide mineral (Wones 1989). The presence of titanite + magnetite + quartz assemblage has long been known as suggestive of relatively high fO 2 in siliceous magmas (e.g., Wones 1989, Enami et al. 1993, and references therein). According to Wones (1989), when this assemblage occurs along with clinopyroxene or amphibole with intermediate or higher Mg# (Mg/(Mg + Fe)) ratios, relatively high fO 2 is implied. Anderson and Smith (1995) suggested that amphibole Fe# ratios for barometry studies should be in the range of 0.40-0.65 (Mg-rich amphiboles). The amphibole from both Major Isidoro and Monteirópolis studied granites present Fe# ratios that vary from 0.577-0.639 and 0.552-0.575, respectively, which are consistent with moderate to high oxygen fugacity (Fig. 9). This exceptional condition remains to be validated (cf. Putirka 2016). Fe-rich amphibole bearing plutons that crystallized under low oxygen fugacity are not suitable for applying hornblende barometry, even though the full mineral assemblage is present and yields high pressures (Anderson and Smith 1995). Fe-rich amphiboles should not have Fe 3+ /(Fe 3+ + Fe 2+ ) ratios lower than 0.20 to apply hornblende barometry (Schmidt 1992, Anderson andSmith 1995). However, exceptions occur when biotite crystallize simultaneously with amphibole and partially incorporate Mg (e.g., Papoutsa and Pe-Piper 2014, Campos et al.  2016). For the studied granites, hornblende Fe 3+ /(Fe 3+ + Fe 2+ ) ratios are within the limiting range except for one core and one rim of the analyzed amphibole from the Monteirópolis granite.
The whole-rock Fe# [Fe/(Fe + Mg)] ratio varies independently of Fe# ratio of amphibole and biotite, and with the increase of fO 2 , the Fe# ratio of these minerals notably decrease (e.g., Anderson andSmith 1995, Anderson et al. 2008). Anderson et al. (2008) proposed an approximate fO 2 relative to the quartz-fayalite-magnetite buffer (Δ QFM ) depending upon Fe# [Fe/(Fe + Mg)] in biotite. For biotite from the Major Isidoro granite, Fe# in biotite ranges from 0.50 to 0.58, and for biotite from the Monteirópolis from 0.43 to 0.55, corresponding to Δ QFM = + 1 to < + 2.5. This result is similar to those for granites in the northern Borborema Province obtained by Campos et al. (2016), who also observed that biotite with high fO 2 (Δ QFM = + 0.8 to + 2.0) crystallizes together with amphibole at similar high oxidizing conditions.
high oxygen fugacity. The pistacite content of epidote of the studied granites (27 to 31mol.%) suggests that fO 2 was higher than QFM and buffered between NNO (nickel-nickel oxide) and HM oxygen buffers (e.g., Sial et al. 2008).

Granitic magma transport estimates
Epidote from the Major Isidoro and Monteirópolis granites is partially included in later minerals such as biotite, plagioclase, and hornblende and shows resorbed rims, especially in contact with quartz and K-feldspar (Fig. 10). Epidote dissolution in residual melts is explained by non-equilibrium conditions when relatively high-pressure epidote is brought up from deeper levels in the magma and mineral aggregates shield and prevent re-equilibration of the epidote inclusions with residual melt (e.g., Brandon et al. 1996). The textural relationships of epidote with other minerals in rocks yield compelling information on upward magma transport (e.g., Brandon et al. 1996, 2008, Long et al. 2005, Brasilino et al. 2011). Sial et al. (2008, based on dissolution kinetic experiments on epidote from Brandon et al. (1996) and estimating the size of partially corroded subhedral epidote crystals, proposed a way to estimate magma transport rates as follows: • Selection of mEp based on their mol.% Ps, and highly corroded subhedral grains which are partially shielded by plagioclase, biotite, or K-feldspar (Fig. 10); • Inferring the original shapes of corroded grains and measure the maximum dissolution zone width; • Estimating the duration of corrosion by using the minimum apparent diffusion coefficient of 5×10 −17 m 2 /s for Si, Al, Ca, and Fe between tonalitic magma and epidote at 750°C (Brandon et al. 1996), as follows: Where: d z = width of dissolution zone (m); D app = apparent diffusion coefficient (5×10 −17 m 2 /s); and t = time for partial dissolution of epidote (s) Accordingly, t = d z 2 /(5×10 −17 m 2 /s). • Depth of host magma emplacement is inferred from Al-in hornblende barometry; • The rate of magma transport is the ratio of the route length (the difference between the emplacement depth and the source depth) to the average time of corrosion of epidote exposed to the host melt. For a tonalite melting, at water-saturated conditions and fO 2 buffered by NNO, plagioclase and epidote may coexist from a depth ~10 kbar (Schmidt and Thompson 1996, Fig. 2). We infer the route length as the difference between 10 kbar and the emplacement depth. Accordingly, T r = L r /t Where: T r = transport rate (m/year); L r = (10 − P e )·10 4 /3 (m); L r = length route (m); P e = pressure of emplacement (kbar); t = time of partial dissolution of epidote (year).
Based on this approach estimated by Sial et al. (2008), we obtain the time for the dissolved width of magmatic epidote and the maximum upward ascension rates of the magmas that formed the Major Isidoro and Monteirópolis granites. This method gives only maximum speed because the calculations assume implicitly that epidote starts reacting with the host magma as pressure drops below 10 kbar assuming, therefore, the minimum possible time for corrosion of epidote crystals. For the Major Isidoro granites, the measured width of dissolution of 0.157 to 0.225 mm occurred during 15.6 to 32 years of partial corrosion, which correspond to ascent rates of 365 to 750 m.year -1 , while for the Monteirópolis granites, the measured dissolution zones of 0.207 to 0.263 mm occurred during 27-49 years of partial corrosion, which correspond to ascent rates of 395 to 635 m.year -1 .
Based on this approach, some limiting assumptions are found as pointed out by Sial et al. (2008): • The error brackets on the Al-in-hornblende barometry, typically of ± 0.6 kbar; • Use of appropriate diffusion coefficients for Ca, Al, Si, and Fe for each magma composition; • Bias in measuring dissolution zone widths of mEp grains (only subhedral grains were used in this study and anhedral grains were ignored); • Decreasing rate of epidote dissolution, leading to an underestimate of digestion time.
Despite these uncertainties, the method provides a good general indication of rapid magma ascent. This is primarily due to the very short calculated dissolution times, which is < 50yr. When compared with ascent timescale from the literature, our rapid rate of upward magma movement suggests diking/conduit flow (10 -1 -10 2 years), and not diapirism (10 6 -10 9 years), as the ascent mechanism (cf. Clemens and Mawer 1992, Petford et al. 1993, 2000. Our calculated dissolution times can be increased by four orders of magnitude and still be short.

Magmatic epidote-bearing granites elsewhere
The occurrence of magmatic Epidote (mEp) and amphibole are key features to characterize the P-T-fO 2 conditions and many textural features have been considered to identify epidote as magmatic (Zen and Hammarstrom 1984, Schmidt and Poli 2004, 2008, such as: • Strong zonation with allanite rich cores; • Inclusion of mEp into biotite, amphibole, and plagioclase; • Poikilitic texture; • Lack of biotite alteration to chlorite; • Lack of plagioclase alteration.

Epidote-bearing granite transport and formation in the Borborema Province
In the Borborema Province, northeastern Brazil, epidote-bearing granites occur commonly throughout Cryogenian-Ediacaran plutons (see Sial and Ferreira 2015). Interestingly, plutons of similar chemical composition, crystallized at similar pressure, may or may not contain magmatic epidote (e.g., Sial et al. 2008). This mineral is highly vulnerable to changing P-T conditions by dissolving during magma ascent; however, the presence of mEp in granitic rocks allows constraining the depth of melting and crystallization, and ascension rates through the continental crust. In this province, magmatic epidote is more abundant and better preserved in older (0.64-0.62 Ga) plutons, however, it occurs in fewer younger (ca. 0.57 Ga) plutons (cf. Ferreira et al. 2004). Ferreira et al. (2011) tied the magmatic epidote survival, which is possible only when upward magma transport from the deep crust is rapid, in the older plutons of the Transversal Domain of the Borborema Province, to hot continental crust conditions during the peak of metamorphism, in the main stage of the Brasiliano orogenic cycle. The Major Isidoro and Monteirópolis plutons are two of these early plutons (~0.63 Ga; Silva et al. 2015Silva et al. , 2016 plausibly produced by partial melting of a reworked Tonian lower-continental crust and by Tonian mantle-derived rocks, respectively, emplaced during the Brasiliano orogenesis along the JHSZ, which must have acted to facilitate the fast upward migration that contributed to prevent mEp dissolution. This is consistent with buoyancy-ascent of magmas in narrow channels, for instance; via an interconnected network of shear zones, as suggested by Collins and Sawyer (1996) and Petford et al. (2000). Narrow conduits (e.g., dikes) ascent has been shown a viable mechanism for the transport of large volumes of calc-alkalic granite melt through the continental crust during a short timescale (Petford et al. 1993).
De Saint Blanquat et al. (2011) found a pluton volume of first-order importance for the duration of pluton construction and observed that the larger a pluton, the longer its construction time. Roughly considering the volume of the studied plutons between 100 and 200 km 3 would give construction time span duration of less than 10 5 years (Petford et al. 2000, De Saint Blanquat et al. 2011, and references therein), which is fairly higher than our fast ascent rates requiring more data (e.g., geophysical to better know the 3-d geometry and accurate volume of the studied pluton) to better understand and discuss. However, this contrast could be partly solved if we consider the growing magmatic pluton source-controlled by magma batches. As showed by Solano et al. (2012), lateral migration of evolved magma through high-porosity melt layers can collect the evolved granitic magma generated from a deep crustal hot zone (DCHZ) of larger areal extent into a localized ascent zone comprising dikes, faults, or shear zones. The segregation of granitic magma from its source is strongly dependent on its physical properties, of which viscosity is the most important one as a function of composition, temperature, and water content (Barker 1998, Giordano et al. 2008, andreferences therein). This property is a critical quantity governing transport ascent in volcanic and magmatic processes. The viscosities of the Major Isidoro and Monteirópolis magmas over their range of composition, water contents, and temperature (59.1-73.4 wt% SiO 2 , 0.36-1.5 wt% H 2 O, and 800-1,000°C, respectively; Tab. 6) were calculated using the model of Giordano et al. (2008), which agreed with subsequent experiments on felsic-mafic magma mixing reality (e.g., Laumonier et al. 2014), providing values of 10 4.1 -10 8.4 , and 10 4.2 -10 8.5 Pa s, respectively (lower than 10 6 Pa s to near-liquidus temperatures of ~1,000°C). Our viscosities values are within those found in the literature for granitic and volcanic magmas that are generally in the range of 10 3 -10 8 Pa s (Scaillet et al. 1998, Clemens and Petford 1999, Laumonier et al. 2014, which Figure 11. Schematic block diagram synthesizing the discussion in the text, from the place of partial melting in a deep crustal hot zone (DCHZ) to the level of emplacement of the Major Isidoro and Monteirópolis plutons (modified from Solano et al. 2012). The drawing is not to scale. After partial melting, the evolved magma may flow laterally and be drained via a network of dikes feeding a main conduit (e.g., dikes, faults, or shear zones) to the plutons (e.g., Brown and Solar 1998, Vanderhaeghe 1999, 2001, 2009, Hall and Kisters 2012. We consider watersaturated conditions and fO 2 buffered by NNO, where plagioclase and epidote may coexist at a depth ~10 kbar. From the ascent route of this depth to the level of emplacement, partial dissolution of epidote occurs during very short calculated dissolution times (< 50yr). evidence very rapid magma ascent rates of up to 1 m.s -1 (e.g., Castro and Dingwell 2009). In Figure 11, we propose a schematic block diagram synthesizing our finds from the place of partial melting in a DCHZ to the level of emplacement of the Major Isidoro and Monteirópolis plutons.

CONCLUSIONS
The Major Isidoro (627 Ma) and Monteirópolis (626 Ma) granites crystallized under oxidizing conditions between the NNO-HM buffers, with initial crystallization temperatures greater than 940°C. Pressures of 5.3-7.7 kbar (Major Isidoro) and 4.3-5.3 kbar (Monterirópolis) were estimated for the final level of emplacement. Epidote survival (partially dissolved) evidences rapid upward transportation rates of the Major Isidoro and Monteiropólis magmas (365-750 m/ year), which were probably favored by the JHSZ, through hot deep to upper continental crust during the onset of the Brasiliano orogeny. These rates are comparable to those from epidote-bearing granites of similar age in other domains of the Borborema Province.

ACKNOWLEDGMENTS
We thank Edward Sawyer for discussions on the geology of the Major Isidoro pluton. Nilson M. Botelho is thanked for mineral chemistry laboratory facilities. We are grateful to Bruna Maria Borba de Carvalho for the acquisition of mineral chemistry at the Electron Microprobe Laboratory in the Universidade de Brasília, Brazil, and Anderson de Abreu da Silva for the XRF chemical analyses at the Stable Isotope Laboratory (LABISE) in the Universidade Federal de Pernambuco, Brazil. Comments on drafts and revised versions of the manuscript by Maia de Lourdes, Anelise Bertoti, Julio Mendes, and Adejardo da Silva Filho, as members of Silva's thesis committee, are acknowledged. We are grateful to Claudio Ricomini, Chief Editor, for careful editorial handling, and especially to Olivier Vanderhaeghe, Silvio Vlach, and an anonymous reviewer for their positive, critical, and constructive comments, which greatly contribute to improve the paper. Obviously, any errors or omissions are solely the responsibility of the authors. VPF and ANS acknowledge the continuous financial support from the CNPq (process number 471034/2012-6) and FACEPE (process APQ-1738-1.07/12). This is the NEG-LABISE contribution n. 293.