5.1. Comparison between new and previous ages from the Quito segment
Overall, our new K-Ar ages are consistent with previous published studies. The Catequilla and Pacpo domes, located northeast and southeast of Casitagua volcano (Fig. 3a), were previously dated between 833 ± 26 and 898 ± 15 ka (Pacheco, 2013; Alvarado et al., 2014) using the same unspiked K-Ar method. We dated two andesitic lava flows sampled on the flanks of Casitagua volcano and obtained close ages of 785 ± 16 and 878 ± 13 ka (20EQ51a and 20EQ52a, respectively; Fig. 3a). These new data suggest that the eruptive activity of Casitagua was contemporaneous with the emplacement of the Catequilla and Pacpo satellite domes. Moreover, this eruptive activity period is stratigraphically consistent with the K-Ar age of 1152 ± 30 ka obtained from the underlying basaltic andesite lavas assigned to the Pisque Formation (Fig. 3a; Alvarado et al., 2014).
In the Inter-Andean Valley, groundmass ages acquired from the Ilaló volcano suggest an activity between 1112 ± 22 and 1273 ± 20 ka (19EQ47 and 19EQ01, respectively; Table 1). This range is younger than the rather poorly constrained whole-rock K-Ar age of 1620 ± 160 ka previously reported (EC 76; Barberi et al., 1988). Although the precise sampling site EC 76 was not provided, it seems to correspond to the lava 19EQ01 located in an easily accessible outcrop on the northwestern flank of Ilaló (Fig. 3a; P. Mothes, pers. Com.). We interpret the age difference due to (1) the presence of phenocrysts within the whole-rock fraction, which may carry inherited argon; and/or (2) the inclusion of weathered grains, for which potassium may have been partially leached out (e.g., Samper et al., 2008; Schaen et al., 2020). Note that such bias towards older ages has already been noted in other lavas in the Ecuadorian arc dated on whole-rock fractions (e.g., Bablon et al., 2018, 2019) and elsewhere (e.g., Quidelleur et al., 1999, 2021; Samper et al., 2008).
5.2. Geochronological data compilation of the Ecuadorian volcanic arc
In order to constrain the onset of the eruptive activity of the current Ecuadorian arc, we have compiled available geochronological data for Ecuadorian volcanoes (Fig. 5). Note that 14C ages are not considered here because, due to their limitation to about 50 ka towards their older bound, they were largely used for the detailed study of the most recent eruptive history of volcanism in Ecuador. Nevertheless, a complilation of these data is offered in Santamaría (2017). Low quality data, such as whole-rock K-Ar ages have been discarded (see section 5.1). Similarly, 40Ar/39Ar ages without minimum quality criteria, such as plateau ages calculated with at least three consecutive steps corresponding to at least 50% of the total 39Ar released, and having an associated isochron with an initial 40Ar/36Ar ratio indistinguishable from the atmospheric value (e.g., Schaen et al., 2020), have also been removed. Finally, following Ramon et al. (2021), we considered volcanic systems to be “potentially active” or “active”, when there was evidence of Holocene or historical activity, respectively (Table 2).
Table 2
Volcanic systems with radiometric ages. Segment refers to Fig. 6. Oldest and youngest ages are given in ka. Volcanoes with a Holocene activity are considered as potentially active (Pot. Active); those with an historic activity or in eruption are considered as active (Bernard and Andrade, 2011); n.d.: not determined or unknown. Dating tech. (dating technique) and minerals used: wr (whole-rock), gms (groundmass), plg (plagioclase), obs (obsidian), FT (fission-tracks). References, 1: Bablon (2018); 2: Beguelin et al. (2015); 3: Bellver Baca et al. (2020); 4: Almeida et al. (2022); 5: Le Pennec et al. (2011); 6: Bablon et al. (2020b); 7: Samaniego et al. (2005); 8: Andrade et al. (2021): 9: Alvarado et al. (2014); 10: Robin et al. (2010); 11: Bigazzi et al. (1992); 12: Bigazzi et al. (2005); 13: Bellot-Gurlet et al. (2008); 14: Hoffer, 2008); 15: Hall et al. (2017); 16: Hidalgo (2006); 17: Santamaria et al. (2023); 18: Santamaria et al. (2022); 19: Hammersley (2003); 20: Opdyke et al. (2006); 21: Bablon et al. (2019); 22: Bablon et al. (2018); 23: Samaniego et al. (2022); 24: Hoffer et al. (2008); 25: Monzier et al. (1999).
Segment | Edifice - Abbreviation | Lat. (°N) | Long. (°E) | Oldest age (ka) | Youngest age (ka) | Dating tech. | Reference |
Tulcan | Chiles - Cerro Negro - CCN | 0.817 | -77.936 | 883 ± 19 | Pot. Active | K-Ar gms | 1 |
Ibarra | Yanaurcu de Piñan - YNP | 0.482 | -78.33 | 3610 ± 60 | 61 ± 20 | 40Ar/39Ar gms | 2 |
Ibarra | Chachimbiro - CCB | 0.467 | -78.312 | 406 ± 20 | Pot. Active | 40Ar/39Ar gms | 3 |
Ibarra | Cotacachi - Cuicocha - CTC | 0.361 | -78.349 | 173 ± 4 | Pot. Active | K-Ar gms | 4 |
Ibarra | Imbabura - IMB | 0.254 | -78.181 | 47 ± 6 | Pot. Active | 40Ar/39Ar wr; K-Ar gms | 5; 6 |
Ibarra | Cubilche - CUB | 0.230 | -78.132 | 45 ± 5 | n.d. | K-Ar gms | 6 |
Ibarra | Cushnirumi - CHS | 0.191 | -78.328 | 411 ± 8 | 383 ± 6 | K-Ar gms | 6 |
Ibarra | Cusín - CUS | 0.159 | -78.148 | 517 ± 8 | 495 ± 12 | K-Ar gms | 6 |
Ibarra | Mojanda - Fuya Fuya - MFF | 0.133 | -78.293 | 1038 ± 87 | 28 ± 5 | K-Ar gms | 6 |
Ibarra | Cayambe - CAY | 0.025 | -77.988 | 409 ± 4 | Active | 40Ar/39Ar wr | 7 |
Quito | Pululahua - PUL | 0.035 | -78.466 | 18 ± 3 | Pot. Active | K-Ar gms | 8 |
Quito | Casitagua - CAS | -0.034 | -78.479 | 1152 ± 30 | 785 ± 16 | K-Ar gms | This study; 9 |
Quito | Pambamarca - PAM | -0.08 | -78.209 | 1374 ± 21 | 1261 ± 18 | K-Ar gms | This study |
Quito | Pichincha - PCH | -0.176 | 78.6 | 1100 ± 10 | Active | 40Ar/39Ar wr | 10 |
Quito | Puntas - PUN | -0.188 | -78.205 | 1132 ± 16 | 1084 ± 17 | K-Ar gms | This study |
Quito | Izambi - IZA | -0.193 | -78.172 | 370 ± 60 | n.d. | FT obs | 11; 12 |
Quito | Coturco - STU | -0.211 | -78.286 | 1959 ± 28 | n.d. | K-Ar gms | This study |
Quito | Chacana - CHN | -0.214 | -78.185 | 2710 ± 190 | Active | FT obs.; 40Ar/39Ar gms | 11; 12; 13 |
Quito | Ilaló - ILA | -0.263 | -78.419 | 1273 ± 20 | 1112 ± 22 | K-Ar gms | This study |
Quito | Pan de Azúcar - PDA | -0.432 | -77.719 | 1150 ± 10 | n.d. | 40Ar/39Ar wr | 14 |
Quito | Antisana - ANT | -0.485 | -78.143 | 378 ± 38 | Pot. Active | 40Ar/39Ar gms | 15 |
Machachi | Atacazo - ATA | -0.357 | -78.619 | 1290 ± 10 | Pot. Active | 40Ar/39Ar gms | 16 |
Machachi | Pasochoa - PAS | -0.467 | -78.481 | 472 ± 8 | 423 ± 10 | K-Ar gms | 17 |
Machachi | Corazón - COR | -0.531 | -78.66 | 178 ± 32 | 67 ± 4 | K-Ar gms | 17 |
Machachi | Sincholagua - SIN | -0.538 | -78.372 | 312 ± 6 | n.d. | K-Ar gms | 17 |
Machachi | Rumiñahui - RUM | -0.581 | -78.507 | 202 ± 8 | n.d. | K-Ar plg | 17 |
Machachi | Almas Santas - ALS | -0.59 | -78.854 | 374 ± 7 | 364 ± 7 | K-Ar gms | 17 |
Machachi | Cosanga - COS | -0.603 | -77.991 | 670 ± 60 | 290 ± 20 | FT obs | 11ou 12 |
Machachi | Santa Cruz - SCR | -0.652 | -78.633 | 702 ± 11 | 60 ± 3 | K-Ar gms | 18 |
Machachi | Iliniza - ILI | -0.663 | -78.716 | 353 ± 6 | Pot. Active | K-Ar gms | 18 |
Machachi | Cotopaxi - COT | -0.681 | -78.438 | 537 ± 11 | Active | K-Ar obs | 17 |
Machachi | Chalupas - CLP | -0.781 | -78.329 | 459 ± 9 | 169 ± 1 | 40Ar/39Ar plg | 19 |
Latacunga | Chinibano - CNB | -0.959 | -78.476 | 1670 ± 190 | n.d. | 40Ar/39Ar gms | 20 |
Latatcunga | Sagoatoa - SAG | -1.155 | -78.669 | 826 ± 12 | 799 ± 12 | K-Ar gms | 21 |
Ambato | Vizcaya - VIZ | -1.333 | -78.427 | 253 ± 5 | n.d. | K-Ar gms | 22 |
Ambato | Puñalica - PLC | -1.399 | -78.678 | 18 ± 3 | n.d. | K-Ar gms | 21 |
Ambato | Huisla - HUI | -1.4 | -78.572 | 612 ± 10 | 492 ± 9 | K-Ar gms | 21 |
Ambato | Carihuairazo - CRH | -1.408 | -78.751 | 512 ± 9 | 157 ± 5 | K-Ar gms | 23; 21 |
Ambato | Mera lavas - MER | -1.408 | -77.927 | 2670 ± 60 | 1920 ± 100 | 40Ar/39Ar wr | 24 |
Ambato | Mulmul - MUL | -1.438 | -78.543 | 174 ± 3 | 145 ± 4 | K-Ar gms | 21 |
Ambato | Puyo cones - PUY | -1.46 | -77.914 | 190 ± 70 | n.d. | 40Ar/39Ar wr | 24 |
Ambato | Chimborazo - CHB | -1.467 | -78.817 | 87 ± 8 | Pot. Active | 40Ar/39Ar gms - K/Ar gms | 23; 21 |
Ambato | Tungurahua - TUN | -1.47 | -78.444 | 293 ± 10 | Active | K-Ar gms | 22 |
Ambato | Igualata - IGL | -1.493 | -78.641 | 376 ± 10 | 107 ± 11 | K-Ar gms | 21 |
Riobamba | Calpi cones - CAL | -1.638 | -78.731 | 62 ± 4 | Pot. Active | K-Ar gms | 21 |
Riobamba | Licto cones | -1.7837 | -78.615 | 183 ± 9 | n.d. | K-Ar gms | 21 |
Riobamba | Sangay - SAN | -2.005 | -78.342 | 380 ± 70 | Active | 40Ar/39Ar gms | 25 |
We have followed the approach considering eruptive subsystems that show a spatial and/or temporal overlap (e.g., Rucu Pichincha and Guagua Pichincha volcanoes) as a single volcano (e.g., Pichincha volcanic complex; Robin et al., 2010). A thorough bibliographic review led to the identification of 77 individual volcanic systems in the Ecuadorian arc (Appendix C), 45 of them with published radiometric ages in the Ecuadorian arc (Table 2). This database includes almost 250 ages (Appendix C), with 145 recent K-Ar ages obtained on groundmass (Fig. 5; Alvarado et al., 2014; Telenchana et al., 2017; Bablon et al., 2018, 2019, 2020a, 2020b; Andrade et al., 2021; Santamaría et al., 2022, 2023; this study; Almeida et al., 2023), 89 40Ar/39Ar ages measured on groundmass or separated crystalline phase (e.g., Monzier et al., 1999; Hammersley, 2003; Samaniego et al., 2005; Hidalgo, 2006; Opdyke et al., 2006; Hoffer, 2008; Hoffer et al., 2008; Hall et al., 2017; Robin et al., 2010; Le Pennec et al., 2011; Bellver-Baca et al., 2020), as well as 15 fission track ages (e.g., Bigazzi et al., 1992, 2005; Bellot-Gurlet et al., 2008).
5.3. Pleistocene eruptive history of the Ecuadorian arc
Figure 6 shows the timing of activity based on the radiometric ages database presented above (Appendix C) and synthetized in Table 2, as well as on stratigraphic and morphological data. Remarkably, the volcanic activity was not homogeneously distributed in space and in time. Three time-intervals of volcanic activity can be identified from Fig. 6 based on the number of active volcanoes, with an early stage older than 1.4 Ma, an intermediate stage between 1.4 and 0.6 Ma, and a late stage for activities younger than 0.6 Ma. Accordingly, the summary of the Pleistocene eruptive history of the Ecuadorian arc is presented below.
5.3.1. Early stage: > 1.4 Ma
The few oldest available ages provide evidence for a volcanic activity occurring from at least 2.5 and up to 1.4 Ma in the Quito and Ambato-Latacunga segments (dark yellow fields, Fig. 7). Note that the occurrence of coeval activity in other segments cannot be ruled out because of the sampling biases towards the most recent units, and the limited number of published studies focusing on the Early Pleistocene volcanism. In addition, the gradual erosion of the oldest deposits and their coverage by younger products prevent us from precisely defining the age of the beginning of the arc construction, although geochronological data seem to point to ~ 2.5 Ma.
Quito segment
40Ar/39Ar plateau ages provided for the Chacana caldera (Western Cordillera) suggest that active volcanism occurred at the center of the present-day Chacana caldera at ~ 2.6 ± 0.2 Ma (Opdyke et al., 2006). We have obtained here a coherent, albeit much more precise, K-Ar age of 2228 ± 34 ka (20EQ85; Table 1) for a basal lava flow forming the southern flank of the caldera. To the east, the construction of Coturco volcano seems to be coeval with this Chacana early stage, as suggested by our K-Ar age of 1959 ± 28 ka (20EQ54; Table 1) acquired from a summit lava. No additional age from this period has been reported for the Western Cordillera or the Sub-Andean zone in the Quito segment.
Coeval activity in the Ambato and Latacunga segments (Back-Arc)
The low-volume mostly effusive volcanism that occurred in the Sub-Andean zone formed various cone-shaped landforms of a few tens of meters height (Cacalurco and Chuvaurcu cones), as well as thin lava sequences to the south of Puyo city (Hoffer et al., 2008; Ball, 2015). A 40Ar/39Ar age of 2770 ± 20 ka was obtained from the northern cone of Cacalurcu volcano (Fig. 7; Hoffer et al., 2008). Note that intense erosion and scarce exposures prevented accurate mapping in this zone. To the north of Puyo, a lava sequence identified as “Mera Lavas” was dated by 40Ar/39Ar at 1980 ± 50 ka (Fig. 7; Hoffer et al., 2008). The source of this lavas remains unknown.
Old ages reported from the Ibarra segment
Four 40Ar/39Ar ages ranging from 3610 ± 60 to 61 ± 20 ka (Béguelin et al., 2015) have been reported for the Yanaurcu de Piñan volcanic system from the Ibarra segment. Such very long activity duration is puzzling as it has not been observed for any other composite volcano of similar size from the whole Ecuadorian arc. Based on their isotopic ratios (Fig. 8 in Béguelin et al., 2015), it can be suggested that the Andesitic Old Yanaurcu formation (~ 3.6 Ma) has a different origin than the younger (< 172 ± 20 ka) products of the Yanaurcu de Piñan volcano. These old products could be related to the Late Miocene Pugarán formation, covering part of the Western Cordillera, and dated at 5.0 ± 2.9 (Boland et al., 2000), or to the Miocene-Pliocene Angochagua formation of the Eastern Cordillera dated between 6.31 ± 0.10 and 3.65 ± 0.07 Ma (Barberi et al., 1988; Boland et al., 2000).
5.3.2. Intermediary stage: 1.4 to 0.6 Ma
During the intermediary construction stage, several new edifices were constructed in the Quito segment and in the southern area of the Ibarra segment (dark to light green fields, Fig. 7). By the end of this stage, new volcanic activity occurred in the segments of Tulcán to the north, and Machachi and Ambato to the south, in which only a few new structures appeared.
Quito segment
In the Eastern Cordillera, the activity of the Chacana caldera (dark green field, 8) is associated with massive lava sequences overlaid by voluminous ignimbrite deposits and obsidian flows (Hall and Mothes, 2008c; Beate and Urquizo, 2015). The sequence cropping out on the western caldera flank was dated between 1350 ± 90 and 1580 ± 70 ka (Bigazzi et al., 2005; Opdyke et al., 2006). To the north, the overlaying El Tablón obsidian flow, dated at 810 ± 50 ka (Bigazzi et al., 2005), provides the younger bound for the Chacana caldera formation. Notably, the eastern flank of the oldest Coturco volcano (Fig. 8) seems to be covered by these sequences, further supporting the new K-Ar age of 1959 ± 28 ka obtained here for this volcano (Table 1).
Further north, the Pambamarca and Puntas volcanoes (Fig. 8) were built on the northwestern and western flanks of Chacana caldera. The eruptive activity of these andesitic volcanoes occurred while Chacana caldera was already active as suggested by our K-Ar ages ranging between 1261 ± 18 and 1374 ± 21 ka for Pambamarca volcano, and between 1132 ± 16 and 1084 ± 17 ka for Puntas volcano (Table 1). The Pambamarca activity ended with the extrusion of the summit rhyolitic dome complex at 1292 ± 20 ka (20EQ62). We propose that the main stage of the Chacana caldera could be more recent or contemporaneous with Pambamarca and Puntas volcanoes, considering that the Chacana ignimbrite series partially covered these edifices (Hall and Mothes, 2008c). The age of the post-caldera activity of Chacana is defined by the Plaza de Armas lava sequence, located south of the caldera (Fig. 3, 8), and dated at ~ 726 ka (Pilicita, 2013), which is roughly coeval to El Tablón obsidian flow, west of the caldera (Ta, Fig. 8), dated at 810 ± 50 ka (Bigazzi et al., 2005).
Volcanic activity in the Western Cordillera seems to have first appear during this stage. In the southern Quito segment, La Carcacha volcano (Atacazo-Ninahuilca volcanic complex) was active at ~ 1300 ka (LC, Fig. 8; Hidalgo, 2006). It was followed to the north by the El Cinto volcano (Pichincha volcanic complex) dated at 1112 ± 24 ka (Fig. 8; Robin et al., 2010). El Cinto edifice was later intruded by the Ungi dome at 910 ± 7 ka (Robin et al., 2010). Further north, the activity of the Casitagua volcano built an andesitic main edifice surrounded by several domes between 898 ± 15 and 785 ± 16 ka (Fig. 8; Alvarado et al., 2014; this study). Then, the construction of the Lower Rucu Pichincha edifice took place between the Casitagua and El Cinto volcanoes. The lava sequences of this volcano were dated between ~ 850 and ~ 590 ka (Robin et al., 2010; Alvarado et al., 2014), and thus are coeval to Casitagua volcano.
In the Guayllabamba valley (Inter-Andean Valley), the activity of Ilaló volcano occurred between 1273 ± 20 and 1112 ± 22 ka forming an edifice composed of andesitic lava sequences (ILA, Fig. 8). Then, in the northern border of the Guayllabamba valley, the emission of voluminous basaltic andesitic lavas were dated at 1152 ± 30 ka (Pisque Formation, L3, Fig. 8; Alvarado et al., 2014) and 1038 ± 87 ka (pre-Mojanda lavas, L2, Fig. 8; Bablon et al., 2020a). Finally, the activity of the Pan de Azúcar volcano in the Sub-Andean zone (Fig. 8) occurred during this stage at 1150 ± 10 ka (Hoffer, 2008).
Extension to the Ibarra and Machachi segments
During this stage, volcanic activity was also present in the Ibarra and Machachi segments (Fig. 8). Viejo Cayambe (Fig. 8) is an ancient edifice formed to the northeast of the Chacana and Pambamarca volcanoes, in the Eastern Cordillera (Ibarra segment). This edifice, dated between 1108 ± 11 and 1050 ± 5 ka, was formed by andesitic to dacitic lava sequences that culminated in a caldera-forming eruption (Samaniego et al., 2005). Further south, the Santa Cruz volcano (Fig. 7) activity occurred in the Inter-Andean Valley ~ 40 km south of La Carcacha volcano (Machachi segment). This edifice was formed by andesitic lava series dated at 702 ± 11 ka, and several dacitic domes in the summit area (Santamaría et al., 2022). Furthermore, several obsidian pebbles were identified in the Cosanga river, in the Sub-Andean zone between the Quito and Machachi segments. These fragments are related to an unknown source located in the Western Cordillera, probably the Cosanga or Bermejo centers (Mothes and Hall, 2008a). Two fission-track ages suggest that part of these obsidians were emitted during eruptions occurring at 670 ± 6 and 290 ± 20 ka (Bellot-Gurlet et al., 2008).
Coeval activity in other segments
Within the Tulcán segment, the lower Cerro Negro edifice (Western Cordillera), located at the Colombian boundary (Fig. 7), was active at 883 ± 19 ka (Telenchana et al., 2017; Bablon, 2018). Further south, in the Latacunga segment, two K-Ar groundmass ages show that the Sagoatoa volcano (Fig. 7) was active from at least 826 ± 12 ka to 799 ± 12 ka (Bablon et al., 2019). We note that two unreliable ages at 1850 ± 190 and 1850 ± 240 ka (Lavenu et al., 1995; Opdyke et al., 2006) were obtained from a sequence of basaltic andesite lavas north of Chinibano volcano (Fig. 7). However, based on the erosional features and by comparison with other eroded volcanoes, we suggest that this volcano could have been constructed during the intermediate stage of the Ecuadorian arc.
5.3.3. Late stage: from 600 ka
During the late construction stage of the Ecuadorian volcanic arc, a striking increase in the number of active volcanoes can be observed in almost all segments (dark and light blue fields, Fig. 9) with at least 50 volcanoes active during this period. Although nearly twenty volcanoes remain poorly studied (most of them from the Tulcán and Latacunga segments), Fig. 9 suggests that most of the volcanic activity in Ecuador extended through time from the Quito segment both northwards and southwards. Considering the available data, it is highly plausible that the eruptive history of most of these poorly studied volcanoes occurred during this stage, i.e., in the last ~ 600 ka. As the eruptive histories of the Ibarra, Machachi, Latacunga, Ambato, and Riobamba segments were presented in details elsewhere (Bablon et al., 2019, 2020a; Santamaría et al., 2023), we only detail here the eruptive history of the Quito segment.
Quito segment
The volcanic activity of the Quito segment was widespread in both Cordilleras and in the Sub-Andean zone. In the Western Cordillera, the eruptive history of the Rucu Pichincha edifice extended to ~ 150 ka, including the sector collapse of its western flank (Robin et al., 2010). The westward migration of the vent caused the formation of the Guagua Pichincha edifice as early as ~ 60 ka, which included two sectoral collapses and the extrusion of several dacitic domes (Robin et al., 2010). The last eruption of Guagua Pichincha took place in AD 1999–2002. Both edifices compose the Pichincha volcanic complex. To the north, the highly explosive activity of the Pululahua dome complex occurred during the Holocene (Andrade et al., 2021). In the Eastern Cordillera, several obsidian and lava flows were emitted during the resurgent phase of the Chacana caldera. The Mullumica and Callejones flows, dated between 200 and 180 ka, are related to major Plinian eruptions whose pumice fallout deposits, known as the Pifo layers, are key stratigraphic markers of the Guayllabamba basin (Bigazzi et al., 2005; Hall and Mothes, 2008c). To the north of the Chacana caldera, the Izambi volcanic series were formed at 370 ± 60 ka (Bigazzi et al., 2005), whereas the underlying lava sequences seems to be related to the activity of Chacana (e.g., Aguilera et al., 2007; Hall and Mothes, 2008c). Unlike the adjacent segments, the volcanic activity in the Inter-Andean Valley appears to be absent during this stage. However, thick volcaniclastic sequences from both Cordilleras were deposited in the Inter-Andean Valley, resulting in the Guayllabamba and Chiche formations lying on top of the lacustrine deposits of the San Miguel formation (Villagómez, 2003; Pacheco et al., 2014). Further west, multiple rhyolitic eruptions occurred on the western edge of the Cordillera Occidental, to the east of the Chacana caldera (e.g., Cosanga and Bermejo centers; Mothes and Hall, 2008a). Although the eruptive histories of these volcanoes remain poorly documented, radiocarbon determinations and stratigraphic relationships suggest that their activity continued into the Holocene (e.g., El Dorado and Pumayacu centers; Hall and Mothes, 2010). Further west, unpublished 40Ar/39Ar ages suggest that the Sumaco volcano (Back-Arc) was active between 255 ± 32 and 200 ± 8 ka (M. Fornari pers. com.), while its most recent activity occurred during the 16th century (Salgado et al., 2021).
Holocene volcanism
From the 77 Quaternary volcanoes identified in our study (Appendix C), 25 were active during the Holocene (Fig. 9). They include Antisana (Hall et al., 2017), Atacazo-Ninahuilca (Hidalgo et al., 2008), Calpi cones (Bablon et al., 2019), Cayambe (Samaniego et al., 1998), Chacana (Hall and Mothes, 2008c), Chachimbiro (Bernard et al., 2014), Chimborazo (Barba et al., 2008), Cotacachi-Cuicocha (Almeida et al., 2023), Cotopaxi (e.g., Hall and Mothes, 2008a), Cubilche (Navarrete et al., 2020), Huañuna (Hall et al., 2017), Iliniza (Santamaría et al., 2022), Imbabura (Le Pennec et al., 2011), Pichincha (Robin et al., 2008), Pululahua (Andrade et al., 2021), Quilotoa (e.g., Di Muro et al., 2008; Mothes and Hall, 2008b), Reventador (e.g., Hall et al., 2004), Sangay (Monzier et al., 1999), Soche (Beate, 1994; Hall and Mothes, 2008b), Sumaco (Salgado et al., 2021), Tungurahua (e.g., Hall et al., 1999; Le Pennec et al., 2013, 2016), and volcanoes located on the western side of the Eastern Cordillera El Dorado, Huevos de Chivo, and Pumayacu (Hall and Mothes, 2010). Although there is no clear evidence of significant explosive activity during the Holocene at the Chiles-Cerro Negro volcano (Santamaría et al., 2017), the unrest episode recorded in AD 2014 (Ebmeier et al., 2016) and present fumarolic activity imply that this volcano may still be potentially active. Finally, due to the lack of more detailed studies, we do not exclude that other volcanoes may have erupted during the Holocene. For instance, the extrusion of the present-day poorly eroded domes of the Mojanda-Fuya Fuya volcanic complex (e.g., Panecillo dome; Robin et al., 2009) and Pambamarca volcano (i.e. Herradura dome; this study), could have occurred during the Holocene.
5.4. Relationship between volcanism and tectonics
The ancient NE-SW crustal structures of the Andean Range (black lines, Fig. 10a; Litherland and Aspden, 1992) seem to partly control the alignments of the Ecuadorian arc, notably for frontal volcanoes. Comparable relationships between the crustal tectonic structures and the arrangement of volcanoes have been also documented in other Andean volcanic zones (e.g., Acocella et al., 2007; Cembrano and Lara, 2009; González et al., 2009; Salas et al., 2017; Tibaldi et al., 2017), and even worldwide (e.g., Takada, 1994; Tibaldi and Bonali, 2017; Bonali et al., 2018). Indeed, numerous studies revealed that the ascension of magma through the crust is mainly controlled by tectonics (e.g., van Wyk de Vries and van Wyk de Vries, 2018). Whereas diapirism is a dominant process in magma transport at deeper levels within the ductile lower crust, dike propagation is prevalent in the upper brittle crust, which is also more prone to brittle tectonics (e.g., Weinberg, 1996; Petford et al., 2000). Thus, multiple authors highlighted the arrangement of eruptive centers and/or dykes as a response to the regional tectonic stress (e.g., Delaney et al., 1986; Cembrano and Lara, 2009; Sychev et al., 2019; Díaz et al., 2020). Moreover, the volcanoes distribution and unrest periods also seem to be affected by local stress regimes and the mechanical properties of the bedrock, which ultimately influence the magma pressure in the conduit (e.g., Gudmundsson, 2006; Gudmundsson and Philipp, 2006; Ebmeier et al., 2016). As a consequence, the spatial distribution of volcanoes can result from both deep (i.e., slab dehydration reactions, mantle heterogeneities) and shallow (i.e., tectonics, structure of the basement) controls. Following the approach of Cembrano and Lara (2009), we discuss below the volcano-tectonic relationship for present-day active and inactive (inherited) crustal structures from the Ecuadorian arc.
Volcanoes over inherited basement structures
In the Western Cordillera, edifices of the Volcanic Front lie above the ancient margin-parallel thrust-belt Pujilí and Toachi sutures that could serve as passive pathways for magma ascent. These NNE-SSW oriented fault systems (black lines, Fig. 10a) separate the Cretaceous oceanic plateau basalts and ophiolites units (e.g., San Juan, Pallatanga, and Pilatón Formations) to the East, from the Paleogene to Eocene volcanic and volcanoclastic units (e.g., Macuchi Formation) to the West (Hughes and Bermúdez, 1997; Hughes and Pilatasig, 2002; Vallejo et al., 2019). The Chachimbiro-Cotacachi-Pichincha-Iliniza-Sagoatoa-Chimborazo volcanoes from the Western Cordillera are located over the Pujilí-Pallatanga suture zone, whereas Almas Santas and Quilotoa, the westernmost volcanoes of the arc, lie over the Chimbo-Toachi suture zone (Fig. 10a; Hughes and Pilatasig, 2002; Vallejo et al., 2019). The relationship between structures of the basement and distribution of the volcanism is less obvious in the Eastern Cordillera. But we note that Pambamarca-Puntas-Sincholagua-Cotopaxi, and Cayambe-Chacana-Antisana-Chalupas eruptive centers display N-S alignments (Fig. 10a). Despite that the reverse faults structures of the western edge of the Eastern Cordillera are hidden by a thick Pliocene volcaniclastic cover (e.g. Pisayambo (Pz) and Angamarca (An) Formations), their existence is inferred by the scarce exposures further north in the Chota Valley, and in the Pisayambo zone to the south (Litherland et al., 1994; Winkler et al., 2005). These major fault systems separate the metamorphic groups formed during the Triassic anatexis and posterior accretion of the Jurassic para-autochthonous terrains (Litherland et al., 1994; Villagómez et al., 2011; Spikings et al., 2015). For instance, the Pambamarca-Puntas-Sincholagua-Cotopaxi volcanoes were constructed over the Peltetec suture zone (Litherland et al., 1994; Vallejo et al., 2019). Further west, the Mojanda-Ilaló-Pasochoa-Rumiñahui alignment might also suggest the presence of an inherited basement structures beneath the Inter-Andean Valley. Nonetheless, exposures of the Inter-Andean valley bedrock are scarce, resulting in an incomplete understanding of the suture architecture created by the accretion of the Western Cordillera terrains (Litherland et al., 1994; Hughes and Pilatasig, 2002). Further east, the Back-Arc volcanoes are built near the basement-involved thrust belt faults related to the Napo Uplift, which was formed by the eastward thrusting of the Western Cordillera (Bès de Berc et al., 2005; Gutiérrez et al., 2019). Although several of these structures are still active, their extension and behavior remain poorly documented.
Volcanoes over present-day compressional or trans-pressional tectonics
Although several active fault systems (fs) have been identified in the Ecuadorian Andes (red lines, Fig. 10b; e.g., Alvarado et al., 2016), their relationship with volcanic landforms is not straightforward (Andrade, 2009). For instance, only a few volcanoes are found near or over the CCPP-fs, the major tectonic structure in Ecuador connecting several trans-pressive and reverse subsystems from the Gulf of Guayaquil (Fig. 1a) to the eastern border of the Eastern Cordillera. The CCPP-fs intersects the Inter-Andean Valley in the Ambato segment, running through the Calpi cones, and Igualata, Huisla, and Mulmul volcanoes (e.g., Bablon et al., 2019; Baize et al., 2020), but does not align with the whole Ambato volcanic cluster. Figure 10b shows that most volcanoes from the arc do not lie above fault-systems, although their evolution might be influenced by fault cinematics (Tibaldi et al., 2007; Alvarado et al., 2016). For instance, the Billecocha subsystem (B-fs) might have influenced the orientation of the Cotacachi sector collapses as well as their subsequent reconstruction patterns (Jomard et al., 2021; Almeida et al., 2023). Further east, the Alambi subsystem (A-fs) played a key role in the preferential growth patterns of the Imbabura volcano (Andrade et al., 2019). The central part of the arc is located in the interaction zone between the Quito and Latacunga reverse fault systems (QL-fs), which are expressed to the north and south of the segment as parallel strands of folds located above major west dipping, blind, en echelon thrust faults (Fiorini and Tibaldi, 2012; Alvarado et al., 2014, 2016). Notably, several segments of the Quito-fs are curved around the volcanic edifices suggesting that the onset of volcanic activity occurred prior to tectonic activity (Alvarado et al., 2014).
5.5. Relationship between volcanism and the Nazca slab and induced crustal stress
The Nazca slab beneath Ecuador has a contorted surface, where the projection of the Grijalva FZ coincides with an inferred slab flexure at depth (Fig. 1a; Yepes et al., 2016). Bablon et al. (2019) proposed that the apparent southward extension of volcanism from the Ambato segment, above the Grijalva FZ, may have been influenced by changes in the slab geometry, and therefore changes in pressure and temperature conditions, during the last 1 Myr. In order to explain the northward spreading of volcanism also observed during this interval, additional mechanisms must be invoked. Unfortunately, the Nazca plate geometry cannot be inferred from seismicity below the central and northern parts of the arc, probably due to the slab thermal conditions (Yepes et al., 2016; Hayes et al., 2018; Araujo et al., 2021). Since at least ~ 2.5 Ma, an intense volcanism seems to be present mainly in the Quito segment, which is above the projection of the Carnegie Ridge. The thermal regime of the subducting young Nazca crust and the Carnegie Ridge thus likely favored magma generation since the Early Pleistocene.
The increase and spreading of volcanism described above may be related to variations in the crustal stress field and/or variations in slab geometry. Indeed, the timing of the increase in the volcanic activity seems to coincide with other regional tectonic events, such as: (1) the cessation of the Late Pliocene WNW–ESE compressional phase registered in the Chota basin (Fig. 2; northern Ecuadorian Andes) prior to the Late Pleistocene, recorded by the deformed dykes and sills (AFT age: 3.7 ± 1.7 Ma) that are unconformably overlain by the undeformed volcaniclastic rocks (ZFT age: 0.5 ± 0.2 Ma; Barragán et al., 1996; Winkler et al., 2005); (2) the formation of the Quito-Latacunga microblock during the Pleistocene as a consequence of the latest transfer stage of the North Andean Sliver deformation zone from the Western Cordillera to the east of the Eastern Cordillera, by creating (or reactivating) and abandoning tectonic structures over time to accommodate convergence directions (Alvarado et al., 2016); and (3) the Late Pleistocene increase of the subsidence and deposition rates reported in the Gulf of Guayaquil (Fig. 1a) recorded by the pile of sediments up to 4 km thick, dated by foraminifera chronobiozones and marine transgression-regressive structures, which overlies the thinner Pliocene series (Witt et al., 2006; Witt and Bourgois, 2010; Loayza et al., 2013; Peuzin et al., 2023). These surface observations of the upper crust dynamics during the Pleistocene might reflect changes at depth, probably due to the increase of coupling between the subducting Carnegie Ridge and the South American plate (e.g., Margirier et al., 2023).