Introduction

Volcanic fields in continental settings commonly consist of scattered or closely spaced small volcanoes, mostly mafic in composition. These volcanoes are often referred to as monogenetic because it is believed that they erupt only once and their activity rapidly wanes (Walker 1993; Connor and Conway 2000; Valentine and Gregg 2008). They commonly form extensive cinder-cone fields, sometimes associated with polygenetic shield volcanoes (Tibaldi 1995; Corazzato and Tibaldi 2006; Dóniz et al. 2008).

The eruptive style of monogenetic basaltic volcanoes can change from Strombolian to Hawaiian to violent Strombolian in response to changing behaviour of the exsolving volatile species, magma rise-speed and the coupling between gas and melt plus crystals (Head and Wilson 1989; Parfitt 2004; Valentine and Gregg 2008; Cimarelli et al. 2010).

Monogenetic volcanism has been related to high regional differential stress, in areas with variable (high or low) magma input/output rate (Takada 1994). Low-output rate volcanic fields usually comprise a small number of widely spaced volcanoes (Valentine and Hirano 2010), while high-output rate volcanic fields comprise dense concentrations of cones, shields and surrounding lava plains (Condit and Connor 1996). The distribution of monogenetic volcanoes within basaltic volcanic fields is generally related to the geometry of the underlying magma source zone (Condit and Connor 1996; Valentine and Keating 2007; Kiyosugi et al. 2010) and/or to the presence of tectonic structures in the shallow crust (Valentine and Perry 2007; Keating et al. 2008).

The Garrotxa Volcanic Field (GVF), the most recent volcanic area in the Iberian Peninsula (0.5–0.01 Ma, Araña et al. 1983; Puiguriguer et al. 2012), consists of more than 40 well-preserved monogenetic centres, grouped in two different zones: (1) a central zone, where the younger cones are located (the central GVF) and (2) a peripheral zone that consists of about 12 widely scattered vents (Araña et al. 1983). Despite their fairly conical morphology, some of these edifices appear to have experienced a complex evolution with alternation of predominantly magmatic activity with minor phreatic and phreatomagmatic episodes (Di Traglia et al. 2009; Gisbert Pinto et al. 2009; Martí et al. 2011). The existing dates for lavas and one paleosol from the central GVF show that the volcanic activity in this sector occurred between 247 ± 17 ka and 11.5 ± 1. ky BP (Guérin et al. 1985; Puiguriguer et al. 2012). Unfortunately, the dated lavas were not placed within a detailed stratigraphic scheme, and this prevents a thorough reconstruction of the spatial–temporal variation of volcanic activity within the central GVF.

We investigated the relationship between eruptive styles and distribution of volcanoes through time within the central Garrotxa tectonic depression.

Detailed stratigraphic analysis and geological mapping of the area (scale 1:5,000) allows us to define multiple volcanic successions and their relative chronology. The facies analysis of the deposits helped us to infer the eruptive styles experienced by each volcanic centre. Volcano morphologies and their internal structures have also been examined by integrating field data and a digital elevation model (DEM) to understand the relation between volcanism and local/regional tectonics. Finally, complementing field data with water well stratigraphy, we were able to reconstruct the top surface of the pre-volcanic basement and to define the thickness and the total volume of the volcanic deposits. Combining the existing geochronology and the volume estimate we have been able to determine the total volcanic output rate in the central GVF.

Geological setting

Geodynamic significance of the North East Volcanic Province of Spain

The GVF is located about 90 km NNE of Barcelona and about 30 km NW of the city of Girona, in the pre-Pyrenees area (Fig. 1a). The GFV together with the Ampurdà (AVF), Tordera (TVF) and Selva (SVF, also called southern Garrotxa; Martí et al. 2011) volcanic fields is part of the North East Volcanic Province of the Iberian Peninsula (NEVP; Araña et al. 1983) also referred as the Catalan Volcanic Zone (Martí et al. 2011).

Fig. 1
figure 1

a North-east Volcanic Province of Spain. AVF: Ampordà Volcanic Field; TVF: Tordera Volcanic Field; SVF: Selva Volcanic Field; GVF: Garrotxa Volcanic Field. Major structural lineaments on land and offshore are reported (after SGC Servei Geològic de Catalunya 1989 and Mauffret et al. 1995). b Schematic geological map of the GVF. Modified after SGC (1989)

Full size image

The Neogene to Quaternary NEVP volcanism is interpreted as the resulting from the Pyrenees post-collisional back–arc extension driven by the African slab retreat, which favoured the rise of magmas through the Iberian continental crust (Gallart et al. 1991; Zeyen et al. 1991; Martí et al. 1992). The NEVP volcanism is associated with a system of NE–SW and NW–SE regional faults that control the geometry of the main tectonic depressions in this area (Araña et al 1983). Volcanic activity in the NEVP took place between 10 and 0.01 Ma, with the oldest eruption in the AVF and the youngest in the GVF (Donville 1976; Araña et al. 1983). Migration of volcanism was coherent with the differential opening velocities of the western Mediterranean (Ligure–Provençal basin) during the Oligocene–Miocene, with its northern sector (Gulf of Lyons) experiencing a faster opening respect to the southern-most part (Valencia margin; Mauffret et al. 1995; Goula et al. 1999). In the same way, volcanism migrated from NE to SW. During the Pliocene–Lower Pleistocene, volcanism migrated to the NW toward the GVF in accordance with the change in the regional stress orientation from pure extensional (vertical σ1 and NW-SE σ3) to mainly strike-slip motions (vertical σ2 and NNW-SSE σ1; Goula et al. 1999). Westward migration of the extensional regime at a rate of 5 mm/year (Vergés and Sàbat 1999) and the consequent migration of volcanism is interpreted in terms of mechanical removal of the mantle lithosphere by buoyancy due to the migration of a thermally heterogeneous mantle (Cebrià et al. 2000). Such buoyancy is inferred to have caused regional scale doming of the crust centred on the tectonically and volcanically active region of the Garrotxa (Lewis et al. 2000).

NEVP parental magmas are petrologically similar to those feeding the Cenozoic to Quaternary intraplate volcanoes of central Europe (Massif Central, Provence Volcanic Field and Eifel Volcanic Field; Mertes and Schmincke 1985; Wilson et al. 1995; Bianchini et al. 2007; Carracedo Sánchez et al. 2012). Volcanic rocks from GVF include leucite basanites, nepheline basanites and alkali olivine basalts.

The pre-volcanic basement

The pre-volcanic basement in the central GVF consists of Paleozoic, Eocene and Quaternary rocks. Paleozoic units underlying the Eocene successions crop out south of the studied area and comprise mainly plutonic and metamorphic rocks (Fig. 1b). The Eocene succession in the Garrotxa area is made up of a deepening–shallowing set of nine depositional sequences broadly represented by alluvial systems capped by delta and shallow carbonate platform systems and finally evolving into a shallowing-upward succession characterised by evaporitic deposits followed by delta and alluvial systems (Gimenez-Monsant et al. 1999). These rocks have been extensively deformed and in the studied area they form E–W-oriented folds dissected by normal faults roughly parallel to their axes (Gallart et al. 1991; Gimenez-Monsant et al. 1999). The Quaternary sedimentary rocks consist of fluvial and lacustrine sediments and limited travertine deposits precedent and contemporaneous to the deposition of the volcanic products (IGC, 2002, 2003).

Two major NW-striking faults, the Amer fault to the west and the Llorà fault to the east, delimit the central Garrotxa tectonic depression. The tectonic depression is further delimited to the north and to the south by the Valfogona thrust and the sedimentary ridge of the Serra Transversal, respectively (Fig. 1b). The GVF is mainly localised in two different sectors of the tectonic depression: the Olot-Fluvià river valley to the north and the Santa Pau-Ser river valley to the south. Some of the major faults controlling the structural architecture of the area proved to be tectonically active in historical times, generating destructive earthquakes during the Middle Ages (Fleta et al. 2001).

The Unconformity Bounded Stratigraphic Units criteria

The fieldwork has been conducted following the Unconformity Bounded Stratigraphic Units criteria (UBSU; Salvador 1987, 1994), based on the hierarchy of the identified stratigraphic unconformities. The reference unit of the UBSU is the Synthem, defined as a stratigraphic body delimited by significant and demonstrable unconformities, preferably having a regional or inter-regional extension (Salvador 1987). Given the volcanic nature of the studied successions, we adopted the UBSU methodology modified for volcanic terrains (de Rita et al. 1997; Di Traglia et al. 2013). Within the framework of the UBSU, the Eruptive Unit (as defined by Fisher and Schmincke 1984) represents the basic lithostratigraphic unit within the Synthem (Fig. 2 for a general example). Different types of unconformities have been identified and classified according to their ranking (Fig. 3). Major unconformities (S1) extend over the whole studied area, characterised by diffuse areal erosion, and are marked locally by decimetre-thick paleosols. The S1 unconformities constitute the boundary of the different Synthems. Lower rank unconformities S2, locally corresponding to paleo-gullies, limit the Eruptive and Epiclastic Units within each Synthem. Other smaller-scale unconformities (S3) are recognised within the Eruptive Units and separate primary (lavas and pyroclastic deposits) and/or syn-eruptive reworked deposits (see Figs. 3 and 4).

Fig. 2
figure 2

Synthetic explanation of the UBSU methodology. The hypothetical cross-section shows the hierarchical relations between unconformity surfaces and the bounded stratigraphic units. A basal unconformity separates the pre-volcanic basement (Prv) from the volcanic deposits. Volcanic deposits are divided into four synthems separated by S1 unconformities (in red). Synthems are formed by primary volcanic deposits (lavas: L; and pyroclastic: P), their syn-eruptive volcaniclastic deposit (Vc) and by the epiclastic deposits (Ep) representing the inter-eruptive sedimentation. Primary and syn-eruptive volcanic deposits form the eruptive units (EU) and are separated by the epiclastic units (Ep) by lower hierarchy S2 unconformities (in blue)

Full size image
Fig. 3
figure 3

Main stratigraphic log correlation of the central area of the GVF. The correlations allowed organising the volcanic successions into four synthems, separated by S1 unconformities of comparable hierarchy (red lines). Each Synthem is composed of one Eruptive Unit, in its turn composed of more than one eruptive episode, and of one Epiclastic Unit (limited by S2 unconformities: blue lines) representing inter-eruptive sedimentation

Full size image
Fig. 4
figure 4

Example of stratigraphic unconformities: a S1 unconformity between Mascou Synthem (RN and LB units) and Les Tries Synthem (CCU, CMU and CBU units). S2 unconformity lies between RN (pyroclastic deposits related to the Roca Negra scoria cone) and LB (epiclastic) units. b S1 unconformities between Mascou Synthem (LB unit) and Les Tries Synthem (CMU, CBU and UB unit). S2 unconformity lies between CBU (syn-eruptive reworking) and UB (epiclastic) units. Several lower-hierarchy unconformities (S3) lie between different pyroclastic deposits (CCU, CMU, CBU) within the same Croscat EU

Full size image

The study of about 50 stratigraphic sections (the more representative of which are sketched in Fig. 4) allowed us to make a geological map of the area (Fig. 5). Classification of the deposits is based on their lithological characteristics, average componentry of clasts and matrix, organisation and geometry and presence of sedimentary structures (Table 1), allowing distinction between primary and secondary deposits (Table 2).

Fig. 5
figure 5

a Geological map of the central GVF area (descriptions and geometry of the Eocene rocks from IGC, 2002, 2003); b legend of the geological map and geochronology from the literature (Guérin et al. 1985). Asterisk refers to 14C dates from Puiguriguer et al. 2012

Full size image
Table 1 Summary of the lithofacies characteristics of the recognised eruptive units
Full size table
Table 2 Summary of the lithofacies characteristics of the reworked deposits
Full size table

Stratigraphy of the Central GVF

Volcanism in the Central GVF produced numerous small monogenetic eruptive centres aligned with the main structural features that controlled the evolution of the Central GVF tectonic depression. Volcanic activity was characterised by both effusive and explosive events that led to the emplacement of lava flows, lava sheets and construction of spatter cones and cinder cones.

Volcanic successions have been organised in four Synthems (Valls dels Arcs Synthem, Santa Pau Synthem, Mascou Synthem, Les Tries Synthem; Fig. 5, Table 1). In each Synthem, one Eruptive Unit (comprising both primary and syn-eruptive reworked deposits containing exclusively volcanic products, Figs. 4, 6 and Table 2) and one Epiclastic Unit (inter-eruptive deposits containing also non-volcanic sediments, Table 2) have been recognised. The three Epiclastic Units represent volcanically quiescent periods. The organisation of the studied successions in Synthems, according to the presence of unconformities of comparable hierarchy, indicates that the volcanism in the central GVF evolved through four main periods during which several eruptions (clusters) occurred, separated by comparable unconformities (i.e. comparable pauses in volcanic activity).

Facies analysis and eruptive style of volcanic centres

Facies analysis of the volcanic deposits allowed us to identify the eruptive styles of the volcanoes, and the transport and sedimentation mechanisms of primary and reworked pyroclastic material. A detailed description of the deposits is given in Table 1. From the facies analysis of the volcanic successions belonging to each eruptive unit, it is possible to reconstruct the eruptive style of each volcanic centre in the central GVF. In general, volcanic activity changed from mainly effusive to prevalently explosive through time. In fact, older eruptive units (Batet de La Serra EU and Puig de Mar EU, Fig. 6) were mainly characterised by the effusion of large lava sheets and lava flows through eruptive fissures. The explosive activity recorded in these early eruptions consisted of localised episodes of Hawaiian-type lava fountaining that constructed spatter ramparts and rootless lava flows. During the intermediate and final eruptive stages (Roca Negra EU and Croscat EU, Fig. 6), in addition to spatter ramparts and lava flows, prolonged Hawaiian and Strombolian activity led to the construction of major, younger, cinder cones. The explosive activity of the youngest Croscat EU culminated with violent Strombolian activity of the Croscat Volcano, which constitutes the most explosive episode known for the central GVF (Cimarelli et al. 2010).

Fig. 6
figure 6

Different facies in primary deposits: a massive to well-laminated, lapilli to ash scoria blanket (Violent Strombolian style, Croscat EU) topped by columnar-jointed lava; b welded bombs and spatter agglomerates of the Puig de Mar EU, produced during Hawaiian-style lava fountaining; c bedded scoria deposits of the Roca Negra EU (Roca Negra scoria cone), produced during Strombolian-type eruptions; d massive to well-bedded, lapilli to ash scoria blanket (Violent Strombolian style, Croscat EU), topped with wavy-bedded scoria-rich sand (post-eruptive reworking) and massive sand with lenses of pebbles from Eocene deposits (inter-eruptive reworking)

Full size image

Cone morphology and morphotectonic features

Methodological approach

Cone morphology and morphotectonic features have been analysed through a dedicated geographic information system (GIS) based on a 5 m resolution DEM. The DEM is derived from the topographic maps of the Institut Cartografic de Catalunya (scale 1:5,000; contour interval 5 m).

The distribution of major morpho-structural features has been deduced by combining analysis of DEM-derived thematic maps such as slope, break in slope and slope aspect (horizontal direction to which the slope faces) with information from the drainage network map, following the methodology described in Cimarelli and de Rita (2006). In order to deduce the main morpho-structural trends of the pre-volcanic basement rocks, this analysis was performed on an area that extends beyond the volcanic deposits. The frequency of the measured lineament azimuth was plotted on rose diagrams (Fig. 7).

Fig. 7
figure 7

DEM-derived maps of morphological elements: a morphotectonic lineaments and related rose diagram (A1); b drainage network elements; rose diagrams of hydrological elements ≥700 m (B1) and ≥800 m (B2); c direction of elongation (yellow) and breaching (red) of cones and craters; d fry plot describing the spatial distribution of the cones and their preferential alignments

Full size image

Information on the orientation of magma-feeding fractures has been obtained for 15 eruptive centres through morphometric investigations (Table 3), following the methodology proposed by Corazzato and Tibaldi (2006). The main morphological parameters used in this analysis are the alignment of the coeval eruptive centres, elongation of the cones, alignment of depressed sectors at the crater rim and the azimuth of breached sectors of the cones. Alignment of coeval cones, cone elongation and alignment of crater-rim depressed points are used as indicators of the direction of the feeding-fracture in the substrate, while cone breaching can either coincide with the weakest zone of the edifice or with the direction of maximum stress applied to its flanks by magma bulging or fracture propagation (Corazzato and Tibaldi 2006; Bonali et al. 2011). Alignment of volcanic centres has been assessed using Fry spatial distribution analysis. This method has been used to determine strain form the distribution of objects in a rock (Fry 1979) and preferred orientation of volcanic cone alignments (Gutiérrez et al. 2005).

Table 3 Morphometric parameters of the volcanic centres
Full size table

Morphological analysis

Our statistical analysis of morphological lineaments in the pre-volcanic basement reveals three main populations of NW-, N- and NE-striking features (Fig. 7a). The most representative NW-striking features are oriented according to the direction of the main transtensional faults limiting the tectonic depression (Amer and Llorát faults). Analysis of the hydrographic features shows that streamlines ≥700 m are scattered in all directions with the main populations represented by E-, NW-, N- and NE-oriented elements (Fig. 7b and B1). With respect to the morpho-structural lineaments of the substrate, the additional E-trending population accounts for the elements incising the volcanic deposits. Here, volcanic deposits masked the underlying morphology of the substrate and forced the drainage system to rearrange accordingly. When restricting the analysis to hydrographic features ≥800 m, the resulting rose-diagram further highlights this effect (Fig. 7, B2).

Investigations of the spatial distribution and morphology of cones provided information on the stress tensor driving the emplacements of feeder dikes and their interaction with pre-existing fractures in the basement. Both alignment of depressed sectors of crater rims and the spatial distribution of eruptive centres reveal NNW- to NW-striking eruptive fractures (Fig. 7c, d; Table 3). Alignments of coeval eruptive centres also show ENE to NE directions, as in the case of the Garrinada-Montsacopa-Montolivet cones. It is worth noting that the azimuths of the breaching sectors are oriented orthogonal to the eruptive fractures determined from the other three morphological parameters, which suggests that cone breaching probably occurred by lava accumulation or bulging of cone flanks (Nolesini et al. 2013).

The volumes of ejecta in edifices have been determined by combining field measurements with the DEM analysis. The volume of each cone has been calculated using the DEM surface enclosed within its basal perimeter (defined using both slope maps and direct field observation), while lava flow volume was calculated by multiplying the flow area by the estimated mean thickness (Table 3). To better constrain changes in flow thickness, several longitudinal profiles were made following the methodology described by Rodriguez-Gonzalez et al. (2010).

Volumes of individual pyroclastic edifices vary from 106 to 108 m3 dense rock equivalent (DRE). Major scoria cones (Croscat, Montsacopa and Garrinada), characterised by larger volumes and a volcanic succession dominated by lapilli and ash, are located near the depocentre of the tectonic depression, while the smaller ones, characterised by lower volumes and coarse-grained materials, are located near the valley margins and are associated with larger volume lava flows.

The Santa Margarida volcano is the only maar-type eruptive centre in the studied area. The thin deposits of this eruptive centre drape the partially excavated Eocene basement and are found intercalated in the eruptive succession of the nearby Croscat Volcano (Martí et al. 2011). As already reported by Di Traglia et al. (2009), the geometry and distribution of the deposits related to the Croscat EU indicate that volcanic activity occurred contemporaneously from multiple vents (Santa Margarida, Croscat, Pomareda and Puig Astrol) along a NNW-striking fracture.

Geometry and volume of the volcanic products

In order to estimate the volume of volcanic material in the central GVF, we modelled the top surface of the Eocene succession based on the stratigraphy of water wells (Custodio Gimena et al. 1984) and field data (Fig. 8). The volume of material between the topographic surface and the top Eocene surface was determined using the GIS by subtracting the elevations of the two surfaces. While the topographic surface is derived from the DEM, the top Eocene surface is interpolated using the stratigraphy of 28 wells intercepting that surface and using the limits of pre-volcanic outcrops as additional surface points (Fig. 8). The volumes of the major cones have also been subtracted, in order to highlight the depocentres in the pre-volcanic basement. Considering the negligible thickness of epiclastic material compared with that of the volcanic products, we estimate the total volume of the volcanics to be about 0.5 km3 DRE.

Fig. 8
figure 8

a Isopach map of the volcanic and volcanoclastic deposits overlying the sedimentary basement. Two major depocentres are visible along the main axis of the Santa Pau-Olot valley. b Simplified geological section of the central GVF area showing the thickness of the volcanic deposits

Full size image

This procedure allowed us to image the geometry of the sedimentary basin and to correlate thickness variations of the deposits with the distribution of the volcanic edifices within the tectonic depression.

The map of Fig. 8a shows two depressions where volcanic products are thicker: One corresponds to Olot village, while the other is located in the central sector of the tectonic depression beneath the Croscat volcano. The Olot village depression, where the Garrinada and Montsacopa scoria cones are located, hosts about 30 m of volcanic material, comprising lavas and subordinate pyroclastic deposits. The second depression (well constrained by water-well log data) extends roughly E-W and hosts about 160 m of volcanic products (Fig. 8b), thus identifying a basin depocentre near the Croscat-Pomareda cones. In both cases, the maximum thickness of volcanic deposits corresponds to the location of the major volcanic edifices of the central GVF.

Discussion

Temporal evolution of volcanic activity

According to the stratigraphy proposed here, each Eruptive Unit represents a series of eruptions clustered in time, occurring between 247 ± 17 and 11.5 ± 1.1 ka (Fig. 4; Guérin et al. 1985; Puiguriguer et al. 2012).

Although we carefully reconstructed the relative chronology of the eruptive products, existing geochronological data do not allow precise determination of the absolute duration of the eruptive and quiescent periods. We can infer that quiescence between the eruptive periods was long enough to allow the formation of the decimetre-thick paleosols and metric erosive surfaces observed in the field (Fig. 3). In particular, given the unfavourable paleoclimatic conditions for mature soil formation, established during the Quaternary in the Pyrenees region (Pallàs et al. 2006), the thickness of the paleosols and the extent of the major erosive surfaces (S1) suggest quite a long interval of inactivity (of the order of tens of kyr based on the available datings; Guérin et al. 1985).

Recent radiocarbon dating of the paleosol contained in the Lower Bosquet Unit (Puiguriguer et al. 2012), coupled with the age (determined by thermoluminescence on plagioclase crystals) of the Upper Santa Pau lava in the upper Roca Negra EU (28.1 ± 2.6 ka; Guérin et al. 1985) indicate quiescence of about 15 kyr between the two last eruptive units. The same calculation can be made for the time interval between Valls dels Arcs and Santa Pau synthems. Considering the age of the Batet de La Serra Lava (247 ± 17 ka, Guérin et al. 1985) and Garrinada Lava (132 ± 12 ka, Guérin et al. 1985), a quiescence of about 115 kyr can be estimated. The dating available for the Lower Santa Pau Lava (Puig de Mar EU) and the Upper Santa Pau Lava (Roca Negra EU) suggest a maximum duration of the Roca Negra EU of about 82 kyr.

Spatial evolution of the volcanic activity

Volcanic activity developed through time with progressive vent migration from the margin toward the centre of the tectonically controlled depression (Fig. 9). This migration is likely to have been favoured by a progressive symmetrical extension of the basin, as indicated by the distribution of maximum thickness of volcanic deposits along the axis of the Olot-Santa Pau valley. Cinder cones in the central GVF are aligned along NNW to NW and ENE to NE directions while morphological characteristics of the cones (cone and crater elongation) prevalently suggest that feeder dykes followed NW to NNW directions. This is in general agreement with the orientation of morphostructural lineaments of the pre-volcanic basement. Magma ascending into the crust is likely to be captured by shallow reactivated faults and fractures, therefore dykes might have near-surface orientations that are not simply perpendicular to the least principal stress direction (Connor and Conway 2000; Valentine and Perry 2007; Le Corvec et al. 2013). Our data show that, in the Central GVF, feeder dikes most probably intruded along pre-existing fractures of the substrate. It is interesting to note that the most recent volcano, Croscat EU, which also represents the peak of explosive activity, has been fed by dikes oriented NNW (Fig. 7), i.e. congruently with the orientation of the present stress tensor in the area (Goula et al. 1999).

Fig. 9
figure 9

Vent migration during emplacement of the four Eruptive Units. The oldest vents (Batet de La Serra EU and Puig de Mar EU) are located near the structural highs, while the youngest (Roca Negra EU and Croscat EU) are located at the centre of the tectonic depression

Full size image

Eruptive style and volcanic output rate

Migration of volcanism toward the centre of the basin coincided with a trend of increasing explosivity, as indicated by the characteristics of the deposits. The occurrence of progressively more explosive activity is characteristic of the volcanoes in the central part of the tectonic depression, where more complex and younger volcanic centres displaying violent Strombolian and phreatomagmatic activity are located. A shift in eruptive style from predominantly effusive and spatter-forming activity in the older and marginal eruptive vents (Batet de La Serra EU and Puig de Mar EU) to more complex and intense activity in the younger and central ones (Roca Negra EU and Croscat EU) is recorded in the stratigraphic succession and facies distribution of the deposits within the eruptive units. Younger eruptions exhibited a broader range of eruptive styles, as occurred during the Croscat eruption with the shift in eruptive intensity from Hawaiian–Strombolian to violent Strombolian (Di Traglia et al. 2009) and the occurrence of a phreatomagmatic episode at the contemporaneously active Santa Margarida volcano (Martí et al. 2011). The dynamics of this eruption, as compared with similar studied eruptions (Valentine et al. 2005; Pioli et al. 2008; Guilbaud et al. 2009) suggests a very high flux of gas-charged magma during this stage of volcanism. Phreatomagmatism occurred in volcanoes located on both high-transmissivity (Montsacopa, Garrinada) and low-transmissivity aquifers (Santa Margarida, Martí et al. 2011), without a specific correlation with aquifer permeability.

In order to estimate the magmatic input rate in the central GVF, we combined volumetric data of emitted products with the tectonic extension rate and the available geochronological datings. We plotted central GVF data (Fig. 10) on the differential-stress versus normalised output-rate diagram proposed by Takada (1994). We used the 0.05–0.125 mm/year vertical slip rate deduced by Fleta et al. (2001) for the Amer fault as an order of magnitude estimate of the differential stress in the studied area during the Quaternary. The results show that the GVF can be classified as a low output rate/small differential stress volcanic field characterised by low average volume of products emitted during each eruptive cluster and by long periods of quiescence between them.

Fig. 10
figure 10

Differential-stress data versus normalised output-rate plot (following Takada 1994). GVF can be classified as a low-output rate, tectonically controlled volcanic field, the activity of which is mainly controlled by tectonics in the NE Catalunya region (stress value from Fleta et al. 2001)

Full size image

Conclusive remarks

We investigated the geological and morphological evolution of the central area of the Garrotxa Volcanic Field and determined the relationship between eruptive styles and volcano distribution through time within the central Garrotxa tectonic depression.

The main outcomes of this study can be summarised as follows:

  • Each recognised synthem bounded by major unconformities (S1) comprises the primary products of volcanic activity and their syn-eruptive reworked deposits (Eruptive Units) and of products of erosion and re-deposition processes (Epiclastic Units). The time period separating eruptive and syn-eruptive deposition from the widespread erosion and deposition of epiclastic material is represented by the S2 unconformity surfaces. The quiescence between two eruptive units is represented in the GVF stratigraphic record by the unconformities produced by erosion or by the sedimentary (epiclastic) deposits, bounded at the bottom and at the top by S2 and S1 unconformities, respectively. The eruptive units represent the products of one or more eruptions clustered in time separated by time-breaks long enough to produce significant unconformities (S3 surfaces).

  • The morphology and alignments of cinder cones in the studied area suggest that feeder dykes followed the orientation of local structural features formed in response to the active stress tensor which characterises the strike-slip tectonic regime active in the NEVP since the Pliocene–Pleistocene.

  • The facies characteristics and distribution of deposits shows that migration of volcanism toward the centre of the basin coincided with a trend of increasing explosivity through time. Phreatomagmatism occurred without a specific geographic correlation and independently from the permeability characteristics of the aquifer underlying the eruptive centres.

  • Given the small average volume of products emitted during each eruptive cluster and the long quiescence between them (15 and 115 kyr), the GVF can be classified as a low output rate/small differential stress volcanic field.

  • The use of the UBSU stratigraphy in volcanic terrains is confirmed as a powerful tool for unravelling the spatial and temporal evolution of volcanic systems. In particular, the versatility of this methodology based on the recognition of surfaces with relative hierarchy allows its application to volcanic areas of different geographic extents. Use of UBSU is particularly effective for basaltic monogenetic volcanic fields in which the homogeneous chemical composition of deposits makes the use of petro-chemical mapping criteria ineffective.

  • In the absence of tephra layers suitable for establishing a tephrostratigraphic framework, correlation of widely extended unconformity surfaces made possible the stratigraphic correlation between widely spaced and low-output rate volcanic centres in this volcanic field.