Unilateral Magma Emplacement of the Telimbela Batholith in the Central Ecuadorian Arc: Implications for Kinematics of Oblique Subduction of the Farallon‐Nazca Plate

The N‐S elongated granodioritic Telimbela batholith (TB, ∼220 km2) was emplaced at the restraining bend of the dextral strike‐slip Calacalí‐Pujilí‐Pallatanga Fault (CPPF) developed at the ca. 120 km wide Tertiary Ecuadorian arc. Our and previous zircon U‐Pb ages indicate that the batholith has a westward younging trend, with the oldest and youngest ages of 25.5 ± 0.5 and 17.5 ± 0.3 Ma, respectively, and a minimum duration of 8 ± 0.6 Ma was acquired. The Al‐in‐hornblende geobarometer results show that the TB was emplaced in the upper crust (∼4.7–5.3 km). The anisotropy of magnetic susceptibility (AMS) results of TB show three distinct groups of magnetic foliations, with a mean orientation of (64°/SE ∠ 39°) (0°/E ∠ 83°), and (16°/NW ∠ 83°) in the eastern, central and western parts of TB, respectively. As the TB was syntectonically emplaced with the evolution of CPPF, the vorticity numbers (Wn) of eastern, central, and western parts are determined by the angle between CPPF and mean orientation of AMS with Wn values of 0.79, 0, and 0.53, respectively. According to previous geodynamic studies on Northern Andes, the early and late kinematic transitions recorded in the TB may correspond to the change of plate subduction direction and the decrease in subduction angle, respectively. Comparing the estimated growth rate of TB to the tempo of Tertiary Ecuadorian arc magmatism revealed by detrital zircon ages and previous plate convergence velocity models, the good consistency among them will help to restore a detailed kinematic evolution of the Ecuadorian arc.

markers of these faults have been mostly obscured and overprinted by later deformation events, and it is a challenge to carry out the kinematic investigation.
In this study, the Late Oligocene-Miocene Telimbela batholith (TB), which is located at the restraining bend of the CPPF (Alvarado et al., 2016), is chosen as the investigation target ( Figure 1b; Schütte et al., 2010b). The N-S elongated TB is parallel with the roughly N-trending CPPF, but the batholith lacks prominent mesoscopic tectonic fabrics. Consequently, we present our field observations, zircon U-Pb age data, geobarometer data, and anisotropy of magnetic susceptibility (AMS) results to reveal the emplacement process of TB and its potential implications for regional tectonics.

Geological Background
Although the accretionary process and subdivision of tectonic units in western Ecuador are still debated, it is generally accepted that the geological framework from west Ecuador has resulted from the eastward subduction of the Farallon-Nazca oceanic plate under the South American plate since the Late Cretaceous ( Figure 1; Vallejo et al., 2009 and references therein). According to Vallejo et al. (2009), Ecuador can be subdivided into five morphotectonic regions from the continental margin to the hinterland, including the coastal lowlands, the Western Cordillera, the Interandean Depression, the Eastern Cordillera and Oriente Basin (Figure 1b). The Tertiary Ecuadorian arc is dominantly developed at the Western Cordillera, which is mainly composed of the Pallatanga and Naranjal blocks and the Macuchi unit that are tectonically juxtaposed and separated by a series of N-to NNE-striking faults (Figure 1b). The Pallatanga block consists of a mantle-derived oceanic plateau extruded at ca. 87 Ma and accreted to the Ecuadorian margin at ca. 75-73 Ma. This block is separated from the Interandean depression, which is covered by thick Pliocene to Pleistocene volcanic rocks, by the dextral strike-slip Calacalí-Pujilí-Pallatanga Fault (CPPF in Figure 1b; Kerr et al., 2002;Spikings & Simpson, 2014). The Naranjal block is situated west of the Chimbo-Toachi Fault (CTF in Figure 1) towards the western Cordillera Occidental margin between 0° and 1°N (Boland et al., 2000;Kerr et al., 2002). This block consists of pillow basalts, basaltic to massive andesitic lavas, and intrusions (Kerr et al., 2002). The Macuchi unit, which is separated from the Pallatanga block by the CTF (Figure 1b; Hughes & Pilatasig, 2002), is mainly composed of basaltic pillow-lavas and breccias, andesitic volcaniclastics high-level intrusions, and lithic tuffs that are covered by the Paleocene to Late Eocene turbiditic and volcaniclastic sediments, as well as shales and cherts (Chiaradia, 2009 and references therein). Previous 40 Ar/ 39 Ar dating on plagioclase in dolerite and andesite from the Macuchi unit yielded plateau ages of 42.62 ± 1.3 and 35.12 ± 1.66 Ma, respectively (Vallejo et al., 2009). The coastal lowlands unit includes a mafic crystalline basement covered by the Paleogene-Neogene forearc deposits (Reynaud et al., 1999).
The TB is a Late Oligocene-Miocene granodioritic intrusion located in the central part of the Ecuadorian arc system (Figures 1 and 2; Schütte et al., 2010aSchütte et al., , 2010b. It is a N-S elongated body with a long axis parallel to the generally N-trending regional structures (Figure 1b). The exposed surface area of TB is up to 220 km 2 with minimum relief of 2.1 km (Figure 2b, elevation data listed in Table S1 in Supporting Information S1). Moreover, the exploration data reveal that the batholith can reach a depth of 3 km. Consequently, the minimum thickness of TB is ca. 5 km. The lithology of TB includes medium-to fine-grained porphyritic granite and medium-to fine-grained granodiorite and diorite ( Figure 2c, Table S1 in Supporting Information S1). Obvious mineral orientation is absent within the pluton, and distinct lithological zonation and intrusive boundaries are not observed in the field due to intensive weathering and vegetation. The pluton contains quartz-diorite dikes in the southernmost part associated with copper mineralization (Schütte et al., 2010b).
The TB intruded into the Paleocene-Eocene Macuchi unit dominated by basalt (Figure 2a). Roof pendants composed of the Macuchi unit are locally preserved in the central part and near the NE margin of the TB,

Zircon U-Pb LA-ICP-MS Dating
A total of six massive granitoid samples (19T04, 19T06, 19T08, 19T09, 19T11, and 19T20) were collected from the TB for zircon U-Pb dating (Figure 2a), and detailed sampling information is provided in Table S2 in Supporting Information S1. Zircon grains were separated and mounted in epoxy resin, and then well-polished for cathodoluminescence imaging. Zircon U-Pb LA-ICP-MS dating was conducted by a Teledyne Photon Machines Analyte He Excimer 193 nm laser ablation system, coupled to an Analytik Jena PlasmaQuant MS Ellite at the Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring of Ministry of Education, Central South University, Changsha, China. The Iolite version 3.71 software was developed by the Isotope Geochemistry Group at the University of Melbourne for U-Pb data processing. A common Pb correction was applied by the standard Microsoft Excel macro (Andersen, 2002). Concordia diagrams and weighted mean calculations were made using Isoplot-R (Vermeesch, 2018). Detailed procedure and parameters of the LA-ICP-MS analysis are enclosed in Text S1 in Supporting Information S1.

Amphibole Composition
Five samples (19T24, 19T14, 19T06, 19T08, and 19T04) were selected for geobarometer measurement according to the age data. The back-scattered-electron images and major composition of amphibole grains were acquired using a JXA-8100 electro microprobe at the Guangzhou Institute of Geochemistry, Chinese Academy of Sciences. Analysis was carried out under an accelerating voltage of 15 kV, a beam current of 30 nA, and a spot size of 5 μm. The analytical errors are usually less than 2%.

Anisotropy of Magnetic Susceptibility (AMS)
Considering the absence of apparently preferred mineral orientation within the TB, AMS analysis was conducted to reveal the cryptic internal fabrics of TB and to shed insight into the magma emplacement process. Oriented samples for AMS analysis were collected using a portable gasoline drill and oriented by both magnetic and solar compasses. The core samples were then cut into 2.2 cm in length and 2.5 cm in diameter in the lab. Due to the study area's intensive weathering and lush vegetation, 20 sample sites are distributed along the major road traverses ( Figure 2a). One hundred fifteen specimens were prepared for AMS measurements (Table S1 in Supporting Information S1). The AMS measurements were performed at the School of Geosciences and Info-physics, Central South University, Changsha, China, using a kappabridge MFK1-FA produced by the AGICO company. The thermomagnetic mineral analysis was conducted at the Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing. Detailed methodology and data processing of AMS refers to Hrouda et al. (2015).

Zircon U-Pb Geochronology
The zircon U-Pb age results are presented as concordia diagrams in Figure 3 and detailed age data are enclosed in Table S2 in Supporting Information S1. Zircon grains from the TB are euhedral, prismatic, and sub-oval, with a length of 200-350 μm (Figures 3h and 3i). Their cathodoluminescence images displayed typical magmatic oscillatory zoning, with no complex internal texture and inherited cores. Their Th/U ratios are higher than 0.31, which is characteristic of magmatic zircons (Table S2 in Supporting Information S1). Zircon U-Pb LA-ICP-MS dating was conducted for six samples. At least 13 spots were analyzed for each sample, and 100 spots were analyzed. The weighted mean 206 Pb/ 238 U ages are adopted as the crystallization ages of the samples.

Emplacement Depth Estimation
Hitherto, dozens of Al-in-hornblende empirical geobarometry equations have been proposed and are widely used to define the emplacement depth of plutons (Erdmann et al., 2014 and references therein). Although the applicability of these geobarometers is still debated, especially for shallow intrusions, strict restrictive conditions of rock compositions can enhance the reliability of the geobarometers. Our microscopic observations indicate that the analyzed amphibole-bearing granitoid samples from the TB contain the required minerals (quartz, alkali feldspar, plagioclase, biotite, hornblende, Fe-Ti oxide, titanite, and apatite) for geobarometry study (Figures 4a and 4b;Mutch et al., 2016). The presence of equilibrium texture (i.e., the triple junctions among polygonal mineral grains, with approximately 120° interfacial angles) in the samples denote was an equilibrium phase achieved in the TB (Figures 4a and 4b). The microprobe analysis results show that the analyzed hornblende grains are magnesiohornblende ( Figure 4c). As the geobarometry equation P (kbar) = 0.5 + 0.331(8) × Al tot + 0.995 (4) × (Al tot ) 2 (with a relative error of ±16% in pressure) proposed by Mutch et al. (2016) is suitable for shallow intrusions (pressure ≥0.5 kbar), it is adopted in this study.
Five amphibole-bearing samples taken from the TB were analyzed, and their barometer calculation results show slightly different average pressure values, for example, 1. 7 ± 0.2 kbar for 19T04, 19T08, and 19T24, 1.9 ± 0.1 kbar for 19T06 and 19T14 (detailed data are listed in Table S3 in Supporting Information S1). A "two-layer" crustal density model of the Macuchi unit was revealed by gravity and aeromagnetic anomaly data. A density of 2.74 and 2.85 g/cm 3 is determined for the upper and lower layers, respectively (Aizprua et al., 2020). In this study, an average density value of 2.80 g/cm 3 is adopted for barometer calculation, and the emplacement depth of the TB is estimated as 4.7-5.4 km.

Magnetic Mineralogy
Microscopic study under reflected light indicates that opaque minerals (magnetite and pyrite) in the granitoids are euhedral to sub-euhedral and occur within or along with phyllosilicate minerals, without obvious clustering (Figures 4a and 4b). Thermomagnetic measurements reveal the variation of magnetic susceptibility (k) with temperature (t) during the heating and cooling processes to better constrain the carrier of susceptibility. Three representative k-t curves of the TB granitoid display a similar weak increasing trend from room temperature to the unblocking temperature of 580°C where the bulk susceptibility drops rapidly ( Figure 5), which suggests that magnetite is the major carrier of susceptibility. These k-t curves are smooth, indicating that the susceptibility is dominated by multi-domain magnetite (Orlický, 1990). Therefore, the three principal vectors of the magnetic ellipsoid are considered as three principal axes of the strain ellipsoid.

Scalar Results of Magnetic Fabrics
Total AMS results are listed in Table S1 in Supporting Information S1. As 85% of the measured specimens have remarkably high mean susceptibility values (K m > 10,000 µSI; Figure 5b), the granitoid in TB are classified as ferromagnetic granitoid according to Bouchez (1997). The corrected degree of anisotropy is low (1.00 ≤ P J ≤ 1.08) to moderate (1.09 ≤ P J ≤ 1.17) (Figure 5b). However, the P J values are not increased with the K m values, suggesting that the degree of anisotropy is not controlled by the intensity of magnetic susceptibility. Furthermore, the shape parameters of 60% and 40% sites are plotted in oblate (0.03 ≤ T ≤ 0.65) and weak prolate (−0.01 ≤ T ≤ −0.53) fields, respectively (Figure 5c).
The AMS results can be subdivided into three groups regarding the zircon U-Pb age results, sampling locations and features of AMS results (Figures 3,  5 and 6). Five early-stage samples that are located in the central-eastern part of the TB (Figures 5 and 6; Table S1 in Supporting Information S1) are characterized by a similar moderate corrected degree of anisotropy (1.05 ≤ P J ≤ 1.07; Figure 5b). The middle-and late-stage samples collected from the central and southwestern parts of the pluton, respectively ( Figure 6 and Table S1 in Supporting Information S1), have a broad spectrum of P J values. Some of them have high P J values (P J > 1.08; Figure 5c), but three middle-stage samples (19T12, 19T13 and 19T14) and two late-stage samples (19T09 and 19T11) that are located close to the large roof pendant or southwestern margin of the batholith have lower P J values (Figures 5 and 6).

Directional Features of Magnetic Fabrics
The equal-area and low hemisphere projections of the AMS results are presented in Figure 6. The AMS results can be subdivided into three fabric groups in terms of the time-spatial distribution and features of site-mean values.
1. The early-stage AMS sites are located in the eastern part of TB and display a relatively consistent orientation of magnetic foliations that trend NE and dip to the SE (green symbols and stereograph in Figure 7a; Table S1 in Supporting Information S1), with a mean magnetic foliation dipping 39° to 54° (great circle in the green stereograph in Figure 7a), except site 19T22 at the margin of the TB that shows a NW-dipping magnetic foliation (small stereograph in Figure 6). Moreover, the inclination is steeper when moving inside the pluton (Figure 7a). The magnetic lineations of these sites plunge to the south or southeast, with variable plunge angles that are increased towards the inside of the batholith (Figures 6 and 7b).   Figure 7a; Table S1 in Supporting Information S1). The relevant magnetic lineations are subhorizontal and subvertical, with a nearly N-S trend (red arrows and stereograph in Figure 7b). 3. The late-stage AMS sites display a steep to subvertical to ENE-to NE-trending, and WNW-to NW-dipping magnetic foliations, with a mean magnetic foliation dipping 83° to 286° (blue symbols and stereograph in Figure 7a; Table S1 in Supporting Information S1). The magnetic lineations plunge dominantly to the SW, with variable plunge angles.

Construction Duration of the TB
Zircon crystallization age generally represents the emplacement age of intrusion in the brittle upper crust. Schütte et al. (2010a) have published a zircon U-Pb age of 25.5 Ma for a granodiorite sample collected from the eastern part of TB, about 2 km from its eastern boundary (asterisk in Figure 2). Our new zircon U-Pb age data display a unilateral westward younging trend (Figures 2  and 3). The youngest zircon U-Pb age in this study is 17.5 Ma, and the relevant sample was collected from the southwestern margin of TB. Therefore, we suggest that the TB was constructed at least from 25.5 to 17.5 Ma, with a minimum construction duration of 8 Ma. Such a long construction duration is comparable to the Tuolumne batholith in Sierra Nevada, North America with an exposure area of ∼1,200 km 2 (approximately six times that of TB) and an inward younging trend .

Formation of AMS in the TB
The high bulk magnetic susceptibility and magnetic mineral analysis results ( Figure 5) indicate that multi-domain magnetite is the main magnetic carrier of granitoid in the TB, and contribution from paramagnetic minerals can be ignored (Bouchez, 1997). Previous studies suggested that the AMS of ferromagnetic granitoids may have resulted from shape anisotropy or the preferred orientation of magnetite (Archanjo et al., 1995;Bouchez, 1997;Tarling & Hrouda, 1993). Our results show no linear correlation between the K m and P J values, and some high K m sites even have low P J values (Figure 5b and Table  S1 in Supporting Information S1), denoting that the degree of anisotropy is not caused by intrinsic anisotropy of magnetite. Furthermore, microscopic observations of the granitoid show that the opaque minerals are usually euhedral and are randomly distributed in or along the contact with phyllosilicates (Figures 4a and 4b), indicating that the AMS is not produced by the cluster of magnetite grains. Field observations show that the TB is not deformed by post-solidification tectonic events. Thus we consider that the magnetic fabrics of TB were produced during the magma emplacement process.

Implications of AMS in the TB
A remarkable feature of AMS in the TB is the subdivision of three distinct magnetic foliation groups that have a mean orientation of (64°/SE ∠ 39°)  (0°/E ∠ 83°), and (16°/NW ∠ 83°) recorded in the eastern, central, and southwestern parts of the batholith, respectively ( Figure 7a). As the construction duration of TB (≥8 Ma) is much longer than the cooling timescale of the magma chamber in the upper crust (Annen et al., 2015), we suggest that the three groups of distinct magnetic foliations were most likely formed during the emplacement of different magma pulses. Furthermore, the N-S elongated pluton geometry (length/width ratio of ca. 3.5/1), the large volume of granitoids (>1,100 km 3 ), and the parallelism/sub-parallelism between the magnetic foliations and regional structures indicate that the TB is a syntectonic pluton. Therefore, the magnetic fabrics in the TB can be used to deduce the regional tectonic regimes. Moreover, as the TB is located at the restraining bend of the crustal dextral strike-slip CPPF fault (Figure 1b; Hughes & Pilatasig, 2002;Kerr et al., 2002), the newly emplaced "softer" TB is more liable to record the regional stress field variation. Thus, we consider that the magnetic fabrics developed in the TB can be used to elucidate the kinematics of the adjacent strike-slip faults.
Previous studies have suggested that the regional kinematic evolution can be revealed through vorticity analysis of syntectonically emplaced plutons (Mamtani, 2014;Mamtani et al., 2013). The vorticity number (W n ) can quantify the degree of non-coaxiality of magma flow (Xypolias, 2010). The flow is usually expressed by particle paths (or streamlines). It is separated into different domains by the "flow apophyses" that do not rotate with respect to instantaneous stretching axes (ISA) during the gradual deformation process (Xypolias, 2010). In terms of the deformation state of particle paths (attract or repulse), the flow apophyses can be recognized as extensional (A e ) or shortening (A S ), and the angle between the two apophyses is α that ranges from 0° to 90° (W n = Cosα; Figure 8a; Passchier, 1986). Mamtani et al. (2013) suggested that the magnetic foliation and shear zone are parallel to the maximum instantaneous stretching axis (ISA max ) and extensional apophysis (A e ), respectively, in a pluton with emplacement affected by the shear zone. The angle between ISA max and A e is ξ, and Weijermars (1991) proposed that W n = Sin2ξ (Figure 8a). The value of W n ranges from 0 to 1, and the W n values ranging from 0 to 0.75 and from 0.75 to 1 correspond to pure shear and simple shear-dominated deformation, respectively (Roy et al., 2016).  The granitoid samples collected from the eastern part of TB contain a NE-trending, early-stage magnetic foliation, with a mean strike of N64°E oblique to the CPPF ( Figure 7a); hence, ξ = 64° and W n = 0.79 (Figures 8b  and 8c). This suggests that the emplacement of early-stage magma was affected by deformation dominated by simple shear as the strike of magnetic foliation is subparallel to the subduction direction, while the magma accretion direction is perpendicular to the strike of magnetic foliation. This suggests that the space for early-stage magma emplacement may have been created by tensional stress perpendicular to the subduction direction, and the tensile fractures may be possible initial space for the early-stage emplaced magma of TB (Figures 8 and 9).
The central part of TB is characterized by a N-striking, middle-stage magnetic foliation (Figure 8), with ξ = 0° and W n = 0. This suggests that the magnetic fabrics of the middle-stage granitoids of TB were developed in the pure shear regime, which is also supported by the absence of obvious ductile shear fabrics inside the TB. This vorticity value denotes that the CPPF was a contractional fault during this period. The vertical to sub-vertical foliation with sub-horizontal lineations also favors this idea (e.g., Garza et al., 2021;Rodriguez-Jimenez et al., 2018). However, the middle-stage granitoid possesses the broadest exposure area but the shortest emplacement time interval among the three stages, which points to a rapid magma ascent and accretion process. As suggested by the qualitative study of Blanquat et al. (1998), magma emplacement in a transpressional zone is driven by tectonic-induced overpressuring that can be transferred into magmatic overpressuring with an upward decreasing pressure gradient. The buoyancy difference within the TB can be ignored due to its similar composition; we thus propose that the magma ascent during the middle-stage was distinctively accelerated by the tectonic-induced overpressuring. Our geobarometer measurements suggest that the TB was emplaced in the brittle upper crust (∼5 km). Therefore, the gravitational collapse in the upper crust will partly offset the compressional strain and facilitate magma emplacement. Moreover, the P J values of the sites close to the pluton roof are lower than those in the deeper part, which also supports this idea (Table S1 in Supporting Information S1).
Magnetic foliations in the western part of TB have a mean strike of N16°E, with ξ = 16° and W n = 0.53 (Figure 8). This indicates that the late-stage magma emplacement was also affected mainly by pure shear in a compressional regime. The late-stage granitoids have similar sub-vertical magnetic foliations and range of P J values to the middle-stage magma phase, which suggests that magma emplacement in both stages occurred in a similar contractional setting.
The probable interpretation of the vorticity variation mentioned above is pivotal for understanding regional kinematic evolution. As the country rocks of TB are dominated by the basaltic Macuchi unit, the vorticity variation caused by rheological contrasts between different lithologies will not be considered here. According to this study, the first change of magnetic foliation strike from N64°E to N-S has occurred near site 19T20 (ca. 23 Ma), that is, the boundary between the early and middle-stage granitoid. This distinct change may be related to the change in subduction direction from NE-SW to roughly E-W (Figure 9; Pilger, 1984). The second change in the strike of magnetic foliations from N-S to N16°E is not prominent and occurs near site 19T04 (ca. 20 Ma) on the boundary between the middle-and late-stage granitoid. However, the plate subduction direction is consistent during these two periods. Subduction of the younger oceanic Nazca plate was initiated at ca. 20 Ma with a lower subduction angle of 25°-35° compared with that of the older Farallon plate (Gutscher et al., 1999;Samaniego et al., 2002;Vallejo, 2007 and references therein). We suggest that the decreased subduction angle will increase the coupling between the amalgamated plates and thus facilitate the partitioning of subduction into the arc-parallel component (i.e., strike-slip component; Figure 9; McCaffrey, 1992).

Estimation of Pluton Width Addition Rate and Its Implications for Regional Geodynamics
It is usually highly challenging to define the pluton growth rate due to erosion or concealment of intrusive bodies as well as limited covering of geochronological data. According to this study, the TB was constructed by a unilateral westward accretion of a series of steep and approximately NNE-and N-trending magma batches. Moreover, the significant composition contrast between the batholith and country rocks denotes that the volume addition of TB caused by assimilation and/ or contamination of the country rocks is minimal due to the thermal limits of the emplaced magma chamber (Glazner, 2007). Consequently, we assume that the width and age interval ratio between two neighboring sampling sites can be regarded as the pluton growth rate. Moreover, the AMS results of sites in the same emplacement stage are characterized by parallel magnetic foliation. Hence, the calculated pluton width addition rate can serve as the average value for that emplacement stage. Detailed calculation results are presented in Table 1.
In particular, sites E07045 and 19T20, which are situated in the early-stage magma emplacement region have a distance of 1.89 km and an age interval of 2.1 Ma (Figure 10a). Therefore, the pluton width addition rate is 0.90 km/Ma, which can be regarded as the average rate of the early-stage magma emplacement process (Figures 10b and Table 1). The distance from site E07045 to the eastern pluton boundary is about 2 km, thus we deduce that the magma emplacement of TB may start at ca. 27.7 Ma (Figure 10a). As the youngest age data (17.5 Ma), was obtained from a sample near the western margin of TB, we consider that the construction duration of TB may be up to 10 Ma.
The middle-stage is the fastest pluton width addition period, with ages ranging from 23.4 Ma (site 19T20) to 21.6 Ma (site 19T06) ( Figure 10). The net pluton width addition of this stage is about 7.03 km (Figure 10a), and thus the pluton growth rate is up to 3.91 km/Ma (Figure 10b). The smallest pluton width addition is 0.14 km between 21.6 Ma (site 19T06) and 20.7 Ma (site 19T04), with the lowest pluton growth rate of 0.16 km/Ma. The growth rate rises to 1.67 km/Ma from site 19T04 (20.7 Ma) to site 19T08 (20.2 Ma) (Figure 10b). We consider that the lowest pluton growth rate period corresponds to the replenishing interval of magma emplacement (Figure 10b). The pluton growth rate of the late-stage period is decreased to 0.58 km/Ma, with ages ranging from 20.2 Ma (site 19T08) to 17.5 Ma (site 19T11).
The pluton growth rate in magmatic arc belts is principally controlled by the magma production-ascent rate and the emplacement space-creating rate, both of which are affected by the plate subduction velocity (Crisp, 1984;Huang & Lundstrom, 2007;Stern, 2020;Yang et al., 2020). Therefore, we consider that the growth rate variation of the TB may have reflected the variation in subduction velocity of the Farallon-Nazca plate, although the subduction velocity is still debated (Figure 10c; Pardo-Casas & Molnar, 1987;Somoza, 1998;Sdrolias & Müller, 2006). Considering the linkage-effect between the plate subduction and the magma emplacement process, we suggest that the pluton growth rate is comparable to the plate convergence rate. The age of initial magma emplacement of the TB is estimated as ca. 27.7 Ma (point ① in Figure 10c), which may correspond to the subduction acceleration at ca. 28 Ma as suggested by Somoza (1998) to account for the incubation time of magma genesis in the source region (Line a in Figure 10b). The fastest pluton growth period of the TB (between point ② and ③ in Figure 10b) is consistent with the peak plate convergence velocity proposed by Somoza (1998). It corresponds to the magmatism flare-up in the Ecuadorian arc from the Late Oligocene to the Miocene (Figure 1c). Furthermore, the first drop of the peak plate convergence velocity in the model proposed by Somoza (1998) occurred at ca. 20 Ma, which coincides with the pluton growth rate drop of the late-stage magma emplacement of the TB (point ④ in Figure 10b).
In the models proposed by Pardo-Casas and Molnar (1987) and Sdrolias and Müller (2006) (Line b and Line c in Figure 10b, respectively), the peak plate convergence velocity began later than 20 Ma (Figure 10b), which corresponds to the late-stage construction of the TB and is contrary to the whole construction process of the TB. Moreover, these two models also conflict with the Ecuadorian arc magmatism tempo since the Eocene, as revealed by the detrital zircon age data (Figure 1c). Our results are more consistent with the plate convergence velocity model proposed by Somoza (1998), whose plate dynamics is more preferable to understand the magmatism in the Ecuadorian arc since the Oligocene time.

Conclusions
The Telimbela batholith was discontinuously intruded into the upper crust (4.7-5.4 km) from the Late Oligocene to the Miocene (27.7-17.5 Ma) with a unilateral westward magma accretion direction. Three distinct groups of magnetic foliations were developed during magma emplacement in the eastern (early-stage), central (middle-stage), and southwestern (late-stage) parts, with a mean orientation of 64°/SE ∠ 39°, 0°/E ∠ 83° and (16°/NW ∠ 83°, respectively. The early-stage NE-striking magnetic foliations developed in a simple shear regime (W n = 0.79) are parallel to the oblique subduction direction of the Farallon plate and perpendicular to the magma accretion trend. Owing to the change in the subduction direction at ca. 23 Ma, the middle-and late-stage N-to NNE-trending magnetic foliations were formed in a pure shear regime, with a W n value of 0 and 0.53, respectively.
The pluton growth rate of TB was estimated based on the zircon U-Pb ages and AMS results. As the emplacement of TB and movement on strike-slip faults in the Tertiary Ecuadorian arc were governed by the oblique subduction of the Farallon-Nazca plate, we consider that the pluton growth rate is comparable to the plate convergence velocity of the Farallon-Nazca plate and that the convergence velocity proposed by Somoza (1998) is preferable to explaining the plate dynamic and Ecuadorian arc magmatism since the Oligocene.

Data Availability Statement
The data used to support the findings of this study are included in the article and in Supporting Information S1 file. Moreover, all the supplementary files were separately uploaded to the Zenodo and can be found online at https://doi.org/10.5281/zenodo.6354896.