Multiple studies have applied zoned garnet geochronology to place temporal constraints on the rates of metamorphism and deformation during orogenesis. We report new high-resolution isotope dilution–thermal ionization mass spectrometry Sm-Nd isochron ages on concentric growth zones from microstructurally and thermodynamically characterized garnets from the Betic Cordillera, southern Spain. Our ages for the garnet core (13.64 ± 0.31 Ma), mantle (13.41 ± 0.37 Ma), and rim (13.34 ± 0.45 Ma) indicate rapid garnet growth and are consistent with published garnet ages interpreted to reflect high-pressure metamorphism in the region. Thermodynamic analysis indicates garnets grew during subduction at ~1.5–2.0 GPa and 570–600 °C. The core to rim duration of spiral garnet growth was just a few hundred thousand years. While other zoned garnet studies have shown similar rapid growth in subduction zone settings, this is the first documentation of such rapid growth of a spiral garnet. Combining this garnet growth duration with the magnitude of spiral inclusion trail curvature, we compute a strain rate of ~10−13 s−1, an order of magnitude faster than all previous spiral garnet studies. We interpret that these spiral garnets recorded a rapid pulse of deformation and strain during the final stages of subduction and incipient exhumation.The abundance of garnet in metamorphic rocks, in combination with its ability to grow in a wide-range of pressure, temperature, and deformational (P-T-D) conditions and to be directly dated, has made it a primary target for investigating tectonometamorphic processes (Baxter and Scherer, 2013). The physical and chemical resistance of garnet to syn- and postgrowth alteration allows it to record and preserve information about its earliest growth stages. Commonly zoned both compositionally and texturally, garnet can provide a near-continuous history of evolving P-T-D conditions. A spectacular example of textural zoning is the development of spiral inclusion trails (Fig. 1A), which form when garnet overgrows an actively evolving matrix foliation during deformation. Spiral trails have been interpreted to form either from syntectonic rotation of the garnet crystal (Rosenfeld, 1968) or from growth over a curved foliation without rotation (Bell, 1985). Here we use “spiral garnet” merely to describe the observed texture with no definitive conclusion toward either model. Either way, spiral garnet provides a critical constraint on the rates of metamorphic mineral growth and deformation during tectonism. Strain rates across orogenic belts have been estimated to be ~10−14 s−1; however, large spatial and temporal variations have been shown, reflecting the localization of strain at different length scales (Pfiffner and Ramsay, 1982; Fagereng and Biggs, 2019). Direct strain rate determinations that combine geochronology and strain data on the same rock are scarce. Christensen et al. (1989, 1994), and Vance and O’Nions (1992) were among the first to employ zoned garnet geochronology to place direct temporal constraints on the rate and duration of garnet growth and link them to the rates of metamorphism and deformation. These early studies reported garnet developing on multi-million-year time scales during regional metamorphism. More recent work on subducted lithologies (Dragovic et al., 2015; Tual et al., 2022) have shown evidence for rapid pulses of garnet growth, on the order of <1 m.y. Prior to our study, Christensen et al. (1989, 1994) and Vance and O’Nions (1992) were the only studies to use zoned garnet geochronology on spiral garnets to calculate a strain rate. Their findings documented strain rates of ~10−14 s−1 during regional metamorphism, consistent with other constraints on regional metamorphic strain rates (Pfiffner and Ramsey, 1982; Müller et al., 2000; Baxter and DePaolo, 2004). To investigate the rates of metamorphism and deformation in the subducted lithologies of the Betic Cordillera, we applied detailed microstructural analysis, thermobarometry, and high-resolution zoned Sm-Nd garnet geochronology to garnets hosting spectacular spiral inclusion trails. The results show very rapid garnet growth and fabric evolution (Fig. 1A).The Betic Cordillera is located at the western termination of the Mediterranean-Alpine orogenic system (Platt et al., 2013). Its metamorphic hinterland, the Alborán Domain, is composed of three nappe complexes, the lowest of which is the Nevado-Filábride Complex (NFC), from which the spiral garnet-bearing samples from this study were collected. The NFC outcrops as a tectonic window in the central and eastern Betics where it forms a broadly east-west–trending anticlinorium including the Sierra Nevada and Sierra de los Filabrides mountains. The NFC has been subdivided into the lower Veleta Complex composed of polymetamorphic Paleozoic graphite schists and quartzites. The overlying Mulhacén Complex (MC) contains similar Paleozoic basement rocks overlain by a Permian–Triassic, and possibly younger, metasedimentary sequences. From bottom to top, they are composed of micaschists, calcschists, and marbles (Puga et al., 2002). Eclogitized mafic and ultramafic lenses within this cover sequence have been interpreted as remains of a narrow branch of the Alpine Tethys (Puga et al., 2011) that originally separated the Iberian paleomargin (including most of the NFC) from a microcontinent that lay to the east and included the overlying Alpujárride and Maláguide Complexes (Platt et al., 2013). The high-pressure lithologies were subducted to pressure of 1.5–2.0 GPa at 20−13 Ma (Sánchez-Vizcaíno et al., 2001; Platt et al., 2006; Kirchner et al., 2016). Exhumation, including low-pressure–high-temperature metamorphic overprinting, commenced as early as 15–13 Ma (Santamaría-López et al., 2019). Recent work by Aerden et al. (2022) illuminated the spiral garnet sample of focus in our study from the MC as a relic from the end of the ca. 13 Ma high-pressure event.Sample 27.1.2 is a metapelite with a mineral assemblage of kyanite (Ky), chloritoid (Ctd), garnet (Gt), phengite (Ph), paragonite (Pg), and quartz (Qtz) with minor amounts of rutile (Ru), ilmenite (Ilm), and apatite (Ap). The abundant centimeter-sized almandine garnets contain abundant inclusions (~50% by volume) of Qtz, Ph, Ctd, and large Ky inclusions within the rims. There is no textural or mineralogical evidence of retrograde overprinting of the stable Ky-Ctd-Gt-Ph-Pg assemblage. Peak pressure-temperature conditions were constrained to ~1.5–2.0 GPa and ~570–600 °C based on the stability of the characteristic high-pressure Gt-Ctd-Ky peak assemblage and the elevated Si concentration of white mica inclusions within garnet and in the matrix (Si = 3.19–3.24 cations per formula unit; Fig. 2). In addition, these peak pressures are consistent with previous estimates for the MC, derived from the associated mafic eclogite bodies (1.5–2.2 GPa; Platt et al., 2006; Puga et al., 2011), from garnet schists of the NFC (2.0–2.2 GPa; Santamaría-López et al., 2019), and for similar Gt-Ctd-Ky schists in the Betics (Smye et al., 2010). Lower pressures (0.4–0.8 GPa) documented in other studies of the NFC (Aerden et al., 2013; Santamaría-López et al., 2019) represent overprinting during exhumation not observed in the rocks of our study. The three-dimensional geometry of the spiral inclusion trails in garnet were characterized using the asymmetry technique of Hayward (1990), with multiple vertical and horizontal cuts of the oriented sample. The inclusion trails spiral up to ~222° with subhorizontal spiral-axes trending predominantly east-west.Preliminary bulk garnet geochronology using isotope dilution–thermal ionization mass spectrometry (ID-TIMS) gave a date of 13.62 ± 0.69 Ma (Aerden et al., 2022). This date agrees with the lower bound of Lu-Hf garnet ages of 18.2–13.3 Ma (Platt et al., 2006) and Rb-Sr multiphase ages of 20.1–13.3 Ma (Kirchner et al., 2016) from the MC. From the same sample, four ~1 cm garnets, with representative inclusion trails, were selected for zoned Sm-Nd geochronology. Using Mn concentration and inclusion trail geometries as proxies for garnet growth, we selected three growth zones for garnets B9, E2, and B8, and two growth zones for garnet D1 (Fig. 1B) for micro-sampling, following the procedures of Pollington and Baxter (2011). Due to small sample size, core and mantle zones from B9, E2, and B8 and rims from B9 and B8 were combined. The six separates were crushed, hand-picked, and sieved to form a coarse garnet separate and a garnet powder separate. The resulting 12 separates were put through a partial dissolution procedure consisting of alternating HF and HNO3 steps to dissolve out contaminating inclusions following the methods of Starr et al. (2020).The eight garnet analyses included on the isochrons (see Table S2 in the Supplemental Material1 for all data) yielded 147Sm/144Nd ratios from 1 to 3.9, indicating the successful removal of problematic inclusions (Baxter and Scherer, 2013). Four garnet analyses were rejected due to low 147Sm/144Nd ratios (<0.3) and poor run behavior. All zones are paired with a whole rock and four matrix separates from which it is assumed the garnets grew in isotopic equilibrium. We include multiple matrix samples to account for possible matrix heterogeneity, which can affect the age (e.g., Gatewood et al., 2015). Figure 3 shows isochrons for each zone analyzed. Sm-Nd isochron dates were calculated using IsoplotR (Vermeesch, 2018), and errors are reported at 2σ. Zones yielded ID-TIMS dates of 13.64 ± 0.31 Ma (cores), 13.41 ± 0.37 Ma (mantles), and 13.34 ± 0.45 Ma (rims).We interpret these new dates to be the ages of core, mantle, and rim growth. Simply taking the difference in age between the core and rim (while ignoring the mantle age) results in a duration of 0.30 ± 0.54 m.y. (2σ). From this, we can constrain the duration of spiral garnet growth to 0.84 m.y. (2σ maximum) to instantaneous. In fact, plotting all data from core, mantle, and rim on a single isochron reveals a combined age of 13.53 ± 0.23 Ma (mean square of weighted deviates = 1.1) which indicates, once again, a very short duration of growth. Alternatively, we can utilize the statistical value of the mantle age and re-compute a duration using a Bayesian statistical model that requires the simple geometric constraint of core to mantle to rim growth (see the Supplemental Material for details). This analysis yields a most probable duration of m.y. 95% credible interval (c.i.). Regardless of the approach taken, our data confirm very rapid growth of these spiral garnets. While other zoned garnet geochronology studies have shown similar rapid rates of garnet growth in subducted terranes (Pollington and Baxter, 2011; Dragovic et al., 2015; Tual et al., 2022), this is the first documentation of such rapid growth in fully developed spiral garnets.Given the high precision of these ages, it is worth considering second-order effects that may alter growth age accuracy, such as diffusional resetting or open system rare earth element (REE) mobility. For garnets of this size (~1 cm) and peak temperatures <600 °C, there will be negligible postgrowth diffusional modification of the outermost rim (e.g., Baxter and Scherer, 2013), which would only serve to make the duration shorter. Open system modification of matrix Sm and Nd content during garnet growth (which could alter apparent ages) is also unlikely given limited solubility of REEs in typical crustal fluids (e.g., Baxter and DePaolo, 2002), lack of evidence for advective flow (i.e., veins), and the fact that all garnet, whole rock, and matrix analyses lie on a tight isochron. These effects do not alter the main conclusion that these spiral garnets grew in just a few hundred thousand years.Whether the observed spiral garnets from the NFC developed by rotation or non-rotation, the development occurred on a rapid time scale spanning just a few hundred thousand years. Christensen et al. (1989, 1994) and Vance and O’Nions (1992) applied zoned garnet geochronology to calculate shear-strain rates from spiral garnet assuming the model of a spherical garnet rotating in a simple shear flow. The shear-strain rates they obtained were × 10−14 s−1, × 10−14 s−1, and 1.9 ± 0.9 × 10−14 s−1, respectively. Adopting the same kinematic model as the previous authors, Biermeier and Stüwe (2003) and Berg et al. (2013) calculated average shear-strain rates of ~6.6 × 10−14 s−1 and ~4.0 × 10−14 s−1, respectively. These two latter studies do not directly date the duration of garnet growth, but rather use thermodynamic modeling under the assumption of reasonable regional heating rates to estimate the duration of garnet growth. Following the same approach with our spiral garnets (with ~222° inclusion trail curvature observed in sections cut normal to the spiral axis) and our modeled garnet growth duration ( m.y.), we obtain a most probable strain rate of × 10−13 s−1 (95% c.i.). This strain rate is an order of magnitude faster than that found in the studies noted above and those determined from methods such as those of Müller et al. (2000) and Pfiffner and Ramsay (1982). Aerden and Ruiz-Fuentes (2020) analyzed spiral garnet from the same outcrop as the garnets described herein and argued that they formed by sequential overgrowth of three crenulation cleavages developed alternately with subvertical and subhorizontal orientations without causing significant garnet rotation (Bell, 1985). This model would also explain why spiral-axes are subparallel to the regional mineral lineation (Aerden et al., 2013), instead of perpendicular as expected in the case of simple shear flow. Berg et al. (2013) also considered a “non-rotational” origin of spiral garnets from the Alps as an alternative possibility to simple shear. Assuming 60% shortening associated with each overgrown crenulation cleavage, Berg et al. (2013) calculated an average strain rate of 8 × 10−15 s−1, which is slower by a factor of five compared to the value they obtained for the case of simple shear (4 × 10−14 s−1). The same approach applied to our NFC spiral garnets, assuming overgrowth of three orthogonal crenulation cleavages, yields a most probable strain rate of × 10−13 s−1 (95% c.i.), which is statistically equivalent to the case of simple shear described above. Regardless of the calculation approach, the spiral garnets in our study record strain rates at least an order of magnitude faster than any other earlier spiral garnet study. Such rapid strain rates have been documented in faults and shear zones, especially over short length scales (Fagereng and Biggs, 2019). Though less likely, our data allow for near instantaneous garnet growth, which would result in strain rates faster than 5.4 × 10−13 s−1 (rotation) and 7.3 × 10−13 s−1 (non-rotation), the upper limit of the 95% c.i.Our thermodynamic modeling suggests that the dated spiral garnets were located at a depth of at least 40 km (~1.5 GPa), as they stopped growing shortly after ca. 13.34 Ma (rim age). While they do overlap statistically, slightly younger allanite U-Pb ages of 12.87 ± 3.31 Ma to 12.40 ± 4.19 Ma obtained by Santamaría-López et al. (2019) from the MC linked to late-stage garnet growth at pressures of 0.7 GPa (~20 km). This implies a minimum exhumation rate on the order of 1–4 cm/yr, similar to exhumation rates deduced in HP-LT rocks by Rubatto and Hermann (2001) and Grujic et al. (2011). The garnet growth in this sample likely spans a rapid shift from burial to exhumation within a localized shear zone.The rapid time scale over which these spiral garnets formed is the novel result of our study. In addition, these spiral garnets record strain rates at least an order magnitude faster than any directly measured by geochronological methods. This is consistent with the conclusion of Fagereng and Biggs (2019) that subduction and the development of young orogenic belts can produce high strains on time scales of several hundreds of thousands of years within faults or shear zones. In the rock we studied, the rapidly growing spiral garnet captures a critical pivot point marking the end of high-pressure metamorphism and shifting rapidly into incipient exhumation.We thank Mike Tappa for help in the clean lab; Mike Mohr for R coding assistance; funding from Spanish grants CGL2015–65602-R (AEI-FEDER), P18-RT-3275, and B-RNM-301-UGR18 (Junta de Andaucía/FEDER); and U.S. National Science Foundation grants PIRE-1545903 and EAR-1946651. In addition, we thank Matthijs Smit, Chris Mark, and an anonymous reviewer for their constructive reviews, and Urs Schaltegger for editorial handling.

中文翻译：

---

高分辨率Sm-Nd石榴石地质年代学揭示俯冲带变质作用期间螺旋石榴石的快速发育

多项研究应用分区石榴石地质年代学对造山作用期间的变质作用和变形速率施加时间限制。我们报告了西班牙南部贝蒂科迪勒拉石榴石的微观结构和热力学特征的同心生长带的新高分辨率同位素稀释-热电离质谱法 Sm-Nd 等时线年龄。我们对石榴石核心 (13.64 ± 0.31 Ma)、地幔 (13.41 ± 0.37 Ma) 和边缘 (13.34 ± 0.45 Ma) 的年龄表明石榴石生长迅速，并且与已发表的石榴石年龄一致，解释为反映该地区的高压变质作用。热力学分析表明石榴石在约 1.5–2.0 GPa 和 570–600 °C 的俯冲过程中生长。螺旋石榴石生长的核心到边缘的持续时间只有几十万年。虽然其他分区石榴石研究也显示了俯冲带环境中类似的快速生长，但这是螺旋石榴石如此快速生长的第一个记录。结合石榴石的生长持续时间和螺旋夹杂物轨迹曲率的大小，我们计算出约 10−13 s−1 的应变率，比之前所有螺旋石榴石研究快一个数量级。我们解释说，这些螺旋石榴石在俯冲的最后阶段和初期折返过程中记录了快速的变形和应变脉冲。变质岩中石榴石的丰度，与其在广泛的压力、温度和温度范围内生长的能力相结合。变形（PTD）条件并可直接测定年代，使其成为研究构造变质过程的主要目标（Baxter 和 Scherer，2013）。石榴石对生长过程中和生长后变化的物理和化学抵抗力使其能够记录和保存有关其最早生长阶段的信息。石榴石通常在成分和结构上进行分区，可以提供近乎连续的 PTD 条件演变历史。结构分区的一个引人注目的例子是螺旋夹杂物轨迹的发展（图 1A），当石榴石在变形过程中过度生长活跃演变的基质叶理时，就会形成螺旋夹杂物轨迹。螺旋轨迹被解释为由石榴石晶体的同构造旋转形成（Rosenfeld，1968）或由不旋转的弯曲叶理生长形成（Bell，1985）。在这里，我们仅使用“螺旋石榴石”来描述观察到的纹理，对任一模型都没有明确的结论。无论哪种方式，螺旋石榴石都对构造运动期间变质矿物的生长和变形速率提供了关键的限制。造山带的应变率估计约为 10−14 s−1；然而，已经显示出较大的空间和时间变化，反映了不同长度尺度上应变的局部化（Pfiffner 和 Ramsay，1982；Fagereng 和 Biggs，2019）。将地质年代学和同一岩石上的应变数据结合起来的直接应变率测定很少。克里斯滕森等人。 （1989、1994）、Vance 和 O'Nions (1992) 是最早采用分区石榴石地质年代学对石榴石生长速率和持续时间进行直接时间限制并将其与变质和变形速率联系起来的人之一。这些早期研究报告称，石榴石在区域变质作用期间在数百万年的时间尺度上发育。最近关于俯冲岩性的研究（Dragovic 等人，2015 年；Tual 等人，2022 年）已经显示出石榴石生长快速脉冲的证据，其数量级约为 <1 在我们的研究之前，Christensen 等人。 (1989, 1994) 以及 Vance 和 O'Nions (1992) 是唯一使用螺旋石榴石分区石榴石地质年代学来计算应变率的研究。他们的发现记录了区域变质作用期间约 10−14 s−1 的应变率，这与区域变质应变率的其他限制一致（Pfiffner 和 Ramsey，1982；Müller 等人，2000；Baxter 和 DePaolo，2004）。为了研究贝蒂科迪勒拉俯冲岩性的变质和变形速率，我们对具有壮观螺旋包裹体痕迹的石榴石应用了详细的微观结构分析、温压测量和高分辨率分区 Sm-Nd 石榴石地质年代学。结果显示石榴石生长和结构演化非常迅速（图1A）。贝蒂科迪勒拉位于地中海-阿尔卑斯造山系统的西端（Platt等，2013）。其变质腹地阿尔博兰域 (Alborán Domain) 由三个推覆杂岩组成，其中最低的是 Nevado-Filábride 杂岩 (NFC)，本研究中的螺旋石榴石样品就是从该杂岩中采集的。 NFC 露头作为贝蒂奇中部和东部的一个构造窗口，在那里形成了一个大致东西走向的背斜层，包括内华达山脉和洛斯费拉布里德山脉。 NFC 已细分为由多变质古生代石墨片岩和石英岩组成的下 Veleta 杂岩体。上覆的 Mulhacén 杂岩 (MC) 包含类似的古生代基底岩石，其上覆盖有二叠纪-三叠纪，以及可能更年轻的变沉积岩层序。从下到上，它们由云母片岩、钙片岩和大理石组成(Puga et al., 2002)。该覆盖层序列中的榴辉岩化镁铁质和超镁铁质透镜体被解释为高山特提斯洋狭窄分支的遗迹（Puga等，2011），该分支最初将伊比利亚古边缘（包括大部分NFC）与位于东部，包括上覆的 Alpujárride 和 Maláguide 综合体（Platt 等，2013）。高压岩性在20~13 Ma时俯冲至1.5~2.0 GPa(Sánchez-Vizcaíno et al., 2001; Platt et al., 2006; Kirchner et al., 2016)。剥露作用，包括低压高温变质叠印，早在 15-13 Ma 就开始了（Santamaría-López 等人，2019）。 Aerden 等人的最新工作。(2022) 阐明了我们研究中焦点的螺旋石榴石样本，它是来自 MC 的一个来自 ca 末端的遗迹。 13 Ma 高压事件。样品 27.1.2 是一种变泥岩，矿物组合为蓝晶石 (Ky)、绿泥石 (Ctd)、石榴石 (Gt)、多硅白硅白云石 (Ph)、钠长石 (Pg) 和石英 (Qtz)，其中少量金红石 (Ru)、钛铁矿 (Ilm) 和磷灰石 (Ap)。丰富的厘米级铁铝榴石石榴石在边缘内含有丰富的 Qtz、Ph、Ctd 内含物（约 50% 体积）和大型 Ky 内含物。没有任何结构或矿物学证据表明稳定的 Ky-Ctd-Gt-Ph-Pg 组合存在逆行叠印。基于特征高压 Gt-Ctd-Ky 峰组合的稳定性以及石榴石和石榴石中白云母包裹体的 Si 浓度升高，峰值压力-温度条件被限制在 ~1.5–2.0 GPa 和 ~570–600 °C。基质（Si = 每个分子式单元 3.19–3.24 个阳离子；图 2）。此外，这些峰值压力与之前对 MC 的估计一致，MC 源自相关的镁铁质榴辉岩体（1.5–2.2 GPa；Platt 等人，2006 年；Puga 等人，2011 年），来自 NFC 的石榴石片岩（ 2.0–2.2 GPa；Santamaría-López 等人，2019），以及 Betics 中类似的 Gt-Ctd-Ky 片岩（Smye 等人，2010）。 NFC 的其他研究中记录的较低压力（0.4-0.8 GPa）（Aerden 等人，2013 年；Santamaría-López 等人，2019 年）代表了在我们研究的岩石中未观察到的折返过程中的叠印。使用 Hayward (1990) 的不对称技术，对定向样品进行多次垂直和水平切割，对石榴石中螺旋夹杂物轨迹的三维几何形状进行了表征。包裹体轨迹螺旋上升至约 222°，近水平螺旋轴主要走向东西向。使用同位素稀释热电离质谱 (ID-TIMS) 进行的初步块状石榴石地质年代学得出的日期为 13.62 ± 0.69 Ma（Aerden 等人，2017）。 ，2022）。该日期与 MC 的 Lu-Hf 石榴石年龄下限 18.2–13.3 Ma（Platt 等人，2006）和 Rb-Sr 多相年龄下限 20.1–13.3 Ma（Kirchner 等人，2016）一致。从同一样品中，选择了四颗具有代表性包裹体轨迹的约 1 厘米石榴石，用于分区 Sm-Nd 地质年代学。使用锰浓度和夹杂物轨迹几何形状作为石榴石生长的代表，我们选择了石榴石 B9、E2 和 B8 的三个生长区，以及石榴石 D1 的两个生长区（图 1B）进行微取样，遵循 Pollington 和巴克斯特（2011）。由于样本量较小，B9、E2 和 B8 的核区和地幔区以及 B9 和 B8 的边缘被合并。将六块分离物粉碎、手工挑选并过筛，形成粗石榴石分离物和石榴石粉末分离物。按照 Starr 等人的方法，对所得 12 个分离物进行部分溶解程序，该程序由交替的 HF 和 HNO3 步骤组成，以溶解出污染夹杂物。 （2020）。对等时线进行的八次石榴石分析（所有数据请参见补充材料1中的表 S2）得出 147Sm/144Nd 比率从 1 到 3.9，表明成功去除了有问题的夹杂物（Baxter 和 Scherer，2013 年）。由于 147Sm/144Nd 比率低 (<0.3) 和运行性能差，四项石榴石分析被拒绝。所有区域都与整个岩石和四个基质分离配对，假设石榴石在同位素平衡下生长。我们包括多个基质样本，以考虑可能影响年龄的基质异质性（例如，Gatewood 等人，2015）。图 3 显示了每个分析区域的等时线。使用 IsoplotR (Vermeesch, 2018) 计算 Sm-Nd 等时线日期，报告误差为 2σ。这些区域得出的 ID-TIMS 日期为 13.64 ± 0.31 Ma（核心）、13.41 ± 0.37 Ma（地幔）和 13.34 ± 0.45 Ma（边缘）。我们将这些新日期解释为核心、地幔和边缘生长的年龄。简单地计算地核和边缘之间的年龄差（同时忽略地幔年龄）会得到 0.30 ± 0.54 my (2σ) 的持续时间。由此，我们可以将螺旋石榴石生长的持续时间限制为瞬时的 0.84 my（2σ 最大值）。事实上，将来自核心、地幔和边缘的所有数据绘制在单个等时线上，显示出总年龄为 13.53 ± 0.23 Ma（加权偏差均方 = 1.1），这再次表明生长持续时间非常短。或者，我们可以利用地幔年龄的统计值，并使用贝叶斯统计模型重新计算持续时间，该模型需要核心到地幔到边缘生长的简单几何约束（有关详细信息，请参阅补充材料）。此分析得出我的 95% 可信区间 (ci) 的最可能持续时间。无论采取何种方法，我们的数据都证实了这些螺旋石榴石的生长速度非常快。虽然其他分区石榴石地质年代学研究表明，俯冲地体中石榴石的生长速度也相似（Pollington 和 Baxter，2011 年；Dragovic 等人，2015 年；Tual 等人，2022 年），但这是首次全面记录这种快速生长情况。开发了螺旋石榴石。鉴于这些年龄的高精度，值得考虑可能改变生长年龄精度的二阶效应，例如扩散重置或开放系统稀土元素（REE）迁移率。对于这种尺寸（~1 cm）和峰值温度<600 °C 的石榴石，最外缘的生长后扩散修改可以忽略不计（例如，Baxter 和 Scherer，2013），这只会使持续时间更短。鉴于稀土元素在典型地壳流体中的溶解度有限（例如，Baxter 和 DePaolo，2002 年），缺乏平流流的证据（即，矿脉），以及所有石榴石、整块岩石和基体分析均位于严格等时线的事实。这些效应并没有改变这些螺旋石榴石在短短几十万年的时间内生长的主要结论。无论是在NFC中观察到的螺旋石榴石是通过旋转还是非旋转形成的，这种发展都发生在短短几百年的快速时间尺度上。一千年。克里斯滕森等人。 (1989, 1994) 以及 Vance 和 O'Nions (1992) 应用分区石榴石地质年代学来计算螺旋石榴石的剪切应变率，假设球形石榴石在简单剪切流中旋转的模型。他们获得的剪切应变率分别为 × 10−14 s−1、× 10−14 s−1 和 1.9 ± 0.9 × 10−14 s−1。采用与之前的作者 Biermeier 和 Stüwe (2003) 以及 Berg 等人相同的运动学模型。 (2013) 计算出的平均剪切应变率分别为~6.6 × 10−14 s−1 和~4.0 × 10−14 s−1。后两项研究并没有直接确定石榴石生长的持续时间，而是在合理的区域加热速率假设下使用热力学模型来估计石榴石生长的持续时间。采用与我们的螺旋石榴石相同的方法（在垂直于螺旋轴切割的截面中观察到约 222° 夹杂轨迹曲率）和我们建模的石榴石生长持续时间 (my)，我们获得了最可能的应变率 × 10−13 s− 1（95% 置信区间）。该应变率比上述研究中发现的应变率和通过 Müller 等人的方法确定的应变率快一个数量级。 (2000) 以及 Pfiffner 和 Ramsay (1982)。 Aerden 和 Ruiz-Fuentes (2020) 分析了与本文所述石榴石来自同一露头的螺旋石榴石，并认为它们是由三个圆齿解理的连续过度生长形成的，这些圆齿解理以近垂直和近水平方向交替发展，而不会引起明显的石榴石旋转 (Bell, 1985)。该模型还可以解释为什么螺旋轴与区域矿物线状近平行（Aerden 等，2013），而不是简单剪切流情况下预期的垂直。伯格等人。 (2013) 还认为阿尔卑斯山螺旋石榴石的“非旋转”起源是简单剪切的替代可能性。 Berg 等人假设每个过度生长的圆齿裂隙都会缩短 60%。 (2013) 计算出平均应变率为 8 × 10−15 s−1，这比他们在简单剪切情况下获得的值 (4 × 10−14 s−1) 慢了五倍。同样的方法应用于我们的 NFC 螺旋石榴石，假设三个正交细圆纹解理过度生长，产生最可能的应变率为 × 10−13 s−1 (95% ci)，这在统计上相当于上述简单剪切的情况。无论采用何种计算方法，我们研究中的螺旋石榴石记录的应变率至少比任何其他早期螺旋石榴石研究快一个数量级。如此快速的应变率已在断层和剪切带中得到记录，尤其是在短长度尺度上（Fagereng 和 Biggs，2019）。虽然可能性较小，我们的数据允许近乎瞬时的石榴石生长，这将导致应变率快于 5.4 × 10−13 s−1 （旋转）和 7.3 × 10−13 s−1 （非旋转），即 95% 的上限我们的热力学模型表明，已测年的螺旋石榴石位于至少 40 公里（~1.5 GPa）的深度，因为它们在大约 10 年后不久就停止生长。 13.34 Ma（边缘年龄）。虽然它们在统计上确实重叠，但 Santamaría-López 等人获得的稍年轻的铜榴石 U-Pb 年龄为 12.87 ± 3.31 Ma 至 12.40 ± 4.19 Ma。 (2019) 来自 MC 与 0.7 GPa（~20 km）压力下的晚期石榴石生长有关。这意味着最小折返率约为 1-4 厘米/年，类似于 Rubatto 和 Hermann (2001) 以及 Grujic 等人在 HP-LT 岩石中推论的折返率。 （2011）。该样本中的石榴石生长可能经历了局部剪切带内从埋藏到掘出的快速转变。这些螺旋石榴石形成的快速时间尺度是我们研究的新结果。此外，这些螺旋石榴石记录的应变率比地质年代学方法直接测量的应变率至少快一个数量级。这与Fagereng和Biggs(2019)的结论一致，即俯冲和年轻造山带的发育可以在断层或剪切带内产生数十万年时间尺度的高应变。在我们研究的岩石中，快速生长的螺旋石榴石占据了一个关键的枢轴点，标志着高压变质作用的结束，并迅速转变为初期折返。我们感谢迈克·塔帕（Mike Tappa）在清洁实验室中提供的帮助； Mike Mohr 提供 R 编码协助；来自西班牙赠款 CGL2015–65602-R (AEI-FEDER)、P18-RT-3275 和 B-RNM-301-UGR18 (Junta de Andaucía/FEDER) 的资金；美国国家科学基金会授予 PIRE-1545903 和 EAR-1946651。此外，我们感谢 Matthijs Smit、Chris Mark 和一位匿名审稿人的建设性评论，并感谢 Urs Schaltegger 的编辑处理。这些螺旋石榴石记录的应变率比地质年代学方法直接测量的应变率至少快一个数量级。这与Fagereng和Biggs(2019)的结论一致，即俯冲和年轻造山带的发育可以在断层或剪切带内产生数十万年时间尺度的高应变。在我们研究的岩石中，快速生长的螺旋石榴石占据了一个关键的枢轴点，标志着高压变质作用的结束，并迅速转变为初期折返。我们感谢迈克·塔帕（Mike Tappa）在清洁实验室中提供的帮助； Mike Mohr 提供 R 编码协助；来自西班牙赠款 CGL2015–65602-R (AEI-FEDER)、P18-RT-3275 和 B-RNM-301-UGR18 (Junta de Andaucía/FEDER) 的资金；美国国家科学基金会授予 PIRE-1545903 和 EAR-1946651。此外，我们感谢 Matthijs Smit、Chris Mark 和一位匿名审稿人的建设性评论，并感谢 Urs Schaltegger 的编辑处理。这些螺旋石榴石记录的应变率比地质年代学方法直接测量的应变率至少快一个数量级。这与Fagereng和Biggs(2019)的结论一致，即俯冲和年轻造山带的发育可以在断层或剪切带内产生数十万年时间尺度的高应变。在我们研究的岩石中，快速生长的螺旋石榴石占据了一个关键的枢轴点，标志着高压变质作用的结束，并迅速转变为初期折返。我们感谢迈克·塔帕（Mike Tappa）在清洁实验室中提供的帮助； Mike Mohr 提供 R 编码协助；来自西班牙赠款 CGL2015–65602-R (AEI-FEDER)、P18-RT-3275 和 B-RNM-301-UGR18 (Junta de Andaucía/FEDER) 的资金；美国国家科学基金会授予 PIRE-1545903 和 EAR-1946651。此外，我们感谢 Matthijs Smit、Chris Mark 和一位匿名审稿人的建设性评论，并感谢 Urs Schaltegger 的编辑处理。