The monogenetic Lemptégy volcano in the Chaîne des Puys (Auvergne, France) was quarried from 1946 to 2007 and offers the possibility to study scoria cone architecture and evolution. This volcano was originally 50–80 m high, but scoria excavation has resulted in a 50-m-deep hole. Beginning in the 1980s, extraction was carried out with the advice of volcanologists so that Lemptégy’s shallow plumbing system and three-dimensional stratigraphy have been preserved. Detailed mapping enabled key stratigraphic units to be distinguished and the constructional phases to be reconstructed. The emplacement and evolution of the shallow plumbing system have also been unraveled. The growth of this monogenetic scoria cone included two temporally well-separated eruptions from closely spaced vents. The activity included Hawaiian, Strombolian and Vulcanian explosions, lava effusion, cryptodome and dome formation, partial collapse, satellite vent formation, eruptive pauses, and intrusion emplacement with consequent uplift. The cone shape, structure, and hence the local stress field, plumbing system, and thermal state were continuously changing, which in turn influenced the eruptive style and location. The plumbing system morphology and microtectonic structures both record local stress field and magmatic flow direction changes. Lemptégy volcano’s internal architecture, stratigraphy, and evolution show how complex a monogenetic volcano can be.
Monogenetic volcanoes are the most prevailing volcanic landform on continents and are mostly mafic in composition (e.g., Wood, 1980; Vespermann and Schmincke, 2000; Valentine and Gregg, 2008). Monogenetic eruptions are common within volcanic cone fields, but also on the flanks of large shield or stratovolcanoes (Davidson and De Silva, 2000; Hintz and Valentine, 2012). Many of these mafic monogenetic volcanoes host basement or mantle xenoliths that are of great interest because they provide a window into the inaccessible mantle and deep crust, as well as providing information on magma origin, ascent rates, and magma-crust interactions (e.g., Rudnick et al., 1993; Jannot et al., 2005; Deegan et al., 2010; Valentine, 2012).
Monogenetic volcanoes are considered to be formed during a single episode of volcanic activity, but field investigations of excavated or eroded volcanoes illustrate that they are the theater of complex interactions with a range of concordant intrusive, effusive, and explosive activity (Connor and Conway, 2000; Valentine et al., 2007; Martin and Németh, 2006; Keating et al., 2008; Riggs and Duffield, 2008; Hintz and Valentine, 2012; Valentine, 2012; Kereszturi et al., 2012). Their structures and growth appear quite simple at first sight, but closer observations reveal many complexities. They can provide evidence of the interplay between regional, magmatic, and volcano-tectonic processes (Riggs and Duffield, 2008; Valentine et al., 2007; Valentine, 2012). In active volcanic fields such as the Chaîne des Puys, future eruptions are likely to occur in locations where no previous cone or dome existed.
Historic eruptions forming monogenetic volcanoes or small scoria cones (e.g., Monte Nuovo, Italy; Cerro Negro, Nicaragua; Parícutin, Mexico; cinder cones on Tolbachik, Kamchatka) have shown that volcanic cones can form over anything from a few days to several years. All such eruptions have had large variations in eruptive style, produced diverse deposits, and have had shifting locations of eruptive vents.
Studies of monogenetic volcanoes have been done within the framework of hazard assessment analysis, looking at the spatial and temporal distribution of volcanoes in monogenetic fields (e.g., Magill et al., 2005; Bebbington and Cronin, 2011; Kiyosugi et al., 2010; Le Corvec et al., 2013), while other studies have been focused on the mechanisms of intrusive and eruptive growth of monogenetic volcanoes, specifically scoria cones (e.g., Valentine and Krogh, 2006; Rapprich et al., 2007; Valentine et al., 2007; Keating et al., 2008; Brenna et al., 2011; Kiyosugi et al., 2012).
The complexity of monogenetic volcano growth can be unraveled through the study of the shallow plumbing complex and deposits. Access to and study of shallow intrusive complexes can be gained indirectly through geophysical and experimental methods or directly on old eroded systems and in quarries (e.g., Williams et al., 1987; Malengreau et al., 1999; Annen et al., 2001; Mathieu et al., 2008; Galland et al., 2009; Delcamp et al., 2012; Gailler and Lénat, 2012; Hintz and Valentine, 2012; Valentine, 2012).
In this study, we perform a detailed survey of the monogenetic mafic scoria cone of Lemptégy, Auvergne, France, which offers an almost complete exposure of the edifice deposit and its underlying shallow plumbing system (Fig. 1A). Lemptégy volcano is part of the monogenetic field of the Chaîne des Puys, and studying this edifice will help to understand the birth and growth of the numerous scoria cones of this world-famous volcanic field. The overall study shows that Lemptégy shares similarities in terms of growth and structure with other small active volcanoes, such as, e.g., Cerro Negro, Nicaragua, and in terms of historic events, Parícutin, Mexico, thus providing insights into the internal processes of active scoria cones. The complex interactions taking place within Lemptégy can occur in the summit regions of larger stratovolcanoes, and so this study also provides information about such larger systems.
According to de Ramond (1815), Lemptégy was already being excavated for its scoria in the early nineteenth century. He noted that Lemptégy was at first considered as a trachyte hill due to its mantling by a trachytic pyroclastic deposit (the Puy Chopine deposit; Boivin et al. 2009a), but the early quarrying uncovered the basaltic scoria interior. Immediately after WWII, excavation began in earnest to provide material for reconstruction, and work continued until 2007. In the late 1980s, the quarrying became increasingly guided by the regular monitoring of volcanologists from the Laboratoire Magmas et Volcans. The shallow plumbing system was progressively uncovered, and the scoria layers were cross sectioned, revealing information about the volcano’s evolution. In 1993, Lemptégy became an educational and tourist attraction, and it now receives over 100,000 visitors per year.
This quarry is a unique place where it is possible to study the first stages of scoria cone construction and the related shallow plumbing complex (Figs. 1B and 1C).
Detailed field work, including stratigraphy, petrography, and microstructure analyses across Lemptégy quarry, was conducted to refine the growth stages, to understand the interactions between the phases of intrusive and extrusive activity, and to document the deformation associated with the volcano growth. Magma flow patterns were also established through meticulous detailing of microstructures, such as tension gashes, striations, and vesicle orientations (Delcamp et al., 2007). These patterns have been confirmed by an anisotropy of magnetic susceptibility (AMS) study (Petronis et al., 2013).
Lemptégy was first described and identified as a volcano in the early nineteenth century (de Ramond, 1815; Scrope, 1858). This volcano belongs to the Chaîne des Puys, a volcanic alignment of around 80 monogenetic and small edifices (Fig. 1A). The Chaîne des Puys has been studied since the early 1800s and was one of the cradles of modern volcanology (Guettard, 1752; Desmarest, 1771, 1773; de Montlosier, 1788; de Dolomieu, 1794, 1798; von Buch, 1819; Scrope, 1858; Lecoq, 1867; Lacroix, 1908). The diversity of edifice morphologies, from scoria cones, lava domes and spines, to maars, makes this volcanic chain a universally representative monogenetic field that has recently been nominated as a United Nations Educational, Scientific, and Cultural Organization (UNESCO) World Heritage site. The diversity of morphologies and eruptive styles reflects the wide span of magma compositions, ranging from basalt to trachyte.
The north-south alignment of the volcanoes is ∼40 km long and runs parallel to the Limagne rift, which is part of the Western European rift (Ziegler, 1994; Merle and Michon, 2001). Around Lemptégy, the Chaîne des Puys splits into several NNE-trending parallel alignments parallel to the Aigueperse fault, which forms part of a the major Rhône-Saône transfer zone between the Limagne and Rhine grabens (van Wyk de Vries et al., 2012). The volcanism in the Chaîne des Puys began to at least 5 m.y. ago, but the present volcanoes are all younger than 200,000 yr, the last eruption being dated at 7600 yr ago (Boivin et al., 2009b).
Lemptégy was a breached scoria cone. The overall history of Lemptégy was established by De Goër de Hervé et al. (1999), who identified two main eruptive centers, Lemptégy 1 and Lemptégy 2 (Figs. 1B and 1C). Lemptégy 1 has been dated at 30 ± 4 ka and Lemptégy 2 at 29.6 ±6 ka by thermoluminescence on lava flows (Guérin, 1983). Stratigraphic relationships confirm Lemptégy 1 as being the older center. Deposits from concurrent eruptions of the neighboring volcano, Puy des Gouttes, are intercalated within the deposits of Lemptégy 1. Deposits from the nearby Puy de Côme (age bracketed between 13 ka and 16 ka; De Goër de Hervé et al., 1999) and Puy Chopine (∼8000 yr B.P.) mantle the Lemptégy 2 cone.
The Supplemental File1 gives a pictorial overview of Lemptégy and is aimed to provide the reader with a clearer view of the overall context of the features described in this study.
Early Lemptégy, referred to as Lemptégy 1, consists of a trachybasaltic cinder cone (Table 1; De Goër de Hervé et al., 1999; Boivin et al., 2009a) and formed a number of vents and lava flows (Figs. 2 and 3). While the topography prior to Lemptégy 1 emplacement has been covered up by numerous eruptions, a drill core at Vulcania Geoscience Park, 500 m to the south, suggests that there was a 100-m-deep valley running just to the south of Lemptégy 1.
The Lemptégy 1 deposit is predominantly an agglomerate deposit, composed of welded or compacted spatter, lapilli, and bombs that are poorly layered or graded, with no clear units being distinguishable. The original height of Lemptégy 1 is estimated at ∼30 m. The deposit is dominated by bombs, which comprise ∼70% of the deposit. The bombs are of various sizes, with a few being 2–3 m wide and many that are meter sized. They are of various shapes, with cowpat, spindle, and bread-crust morphologies. The presence of lava ball bombs, welded facies, red-dominated deposits, and meter-sized bombs indicates near-vent or within-vent deposits.
The lower 5 m section of the Lemptégy 1 deposit is strongly compacted and cemented and is dark gray in color, while the upper part, which forms the majority of the outcrop, is red. There is a fine, scoriaceous lapilli element within the agglomerate that becomes dominant toward the top, where it replaces the agglomerates, in the form of an overlying tephra layer. This tephra blankets all the Lemptégy 1 deposits, except over an uplifted area, where it is has slumped off. Those scoriaceous lapilli are interpreted as being from the Puy de Gouttes volcano (De Goër de Hervé et al., 1999).
The deposits are crisscrossed by faults that have either brecciated the clasts or have created dilatant zones where voids have opened up. These latter are now being preferentially eroded, and a conjugate pattern of faults can be observed (Fig. 2). The faults with largest throws originate from dikes that pushed and uplifted the pyroclastic layers.
Two pahoehoe lava flows are seen in the northeast side of the Lemptégy 1 deposit (Figs. 2 and 4A). The zone below the lavas is a 4-m-high bulbous mass of highly vesiculated rock that narrows downward to a 40-cm-wide dike, oriented N150° (dike D4). The sides of the bulbous mass are composed of intensely crushed agglomerate. The dike has an intensively sheared margin and connects southward into a strongly sheared fault plane, cutting the Lemptégy 1 agglomerate (Fig. 4A). The pahoehoe lava flows, bulbous mass, and the dike (dike 4) were seen to be connected during the excavation, and we interpret the ensemble as a feeding dike, vent, and lava flows. The two pahoehoe units are intercalated with the Puy Des Gouttes scoria deposit, and we thus correlate them temporally with the late stage Puy de Gouttes activity after the initial construction of Lemptégy 1. The lavas are not brecciated or fractured, so they must have been erupted after the deformation of the adjacent faulted zone.
One small area at the south side of the Lemptégy 1 deposits is a massive lava (Fig. 3), in which tension cracks and shear planes are frequently visible. The distribution indicates flow down toward the southeast, and we interpret this feature as a clastogenic lava flow within the agglomerates.
A clear unconformity is observed between the Lemptégy 1 deposits and those associated with Lemptégy 2 (Fig. 4B). The nature of the unconformity varies across the quarry and occurs as a centimeter-thick weathered horizon or a centimeter-thick soil, or as slightly discolored Puy des Gouttes deposits between the two eruptive deposits.
Shallow Plumbing System
The excavated Lemptégy 1 shallow plumbing system includes four main dikes exposed in the quarry (Figs. 2 and 3). Dike 1 runs close to the later Lemptégy 2 center, and a depression can be seen in the Puy de Gouttes and Lemptégy 1 scoria above the dike tip, indicating that it was a late-intruded dike. Dike 1 connects laterally and upward with dike 2, into a central dike, dike 3 (Fig. 4C). The fourth dike is set apart from the others and (dike 4) can be traced to the two pahoehoe lavas mentioned above (Fig. 4A).
Dike 3 shows shear features on its flank, such as striae, indicating that it was injected laterally into a fault and/or created a fault ahead of its propagating tip (e.g., Mathieu et al., 2008). The fault, oriented N000°, branches and splays upward into a strike-slip flower structure. This flower structure is one of the major features shown to visitors, as it provides an exceptional demonstration of a horst and graben system (Figs. 2, 4B, and 5). This major N000° fault has a vertical dip, and it is intimately connected to the sheared dike 3, which shares this orientation, cuts across the volcano, and has striations indicating strike-slip movement (Figs. 2 and 4C). The tension gashes and opening cracks along this fault and the associated dike show in general that the movement was sinistral. In line with this fault, there are open fractures in the Lemptégy 1 and Lemptégy 2 lava flows in the Vulcania Geoscience Park (Fig. 1).
Because the dikes cut, or cause deformation of, all the overlying Puy de Gouttes strata, they must have been intruded very late during the Lemptégy 1 history (Fig. 5). For example, dike 1 has a depression that is filled by Puy de Gouttes tephra, dike 3 is connected via the strike-slip flower and graben structure to grabens partly filled by Puy des Gouttes deposits, and dike 4 has fed the two late pahoehoe lavas. The dikes are thus associated with the uplift and deformation of the Lemptégy 1 deposit toward the end of the Puy des Gouttes–Lemptégy 1 eruption. As previously described, Puy Des Gouttes tephras cover all the Lemptégy 1 deposit, except over one area, where the Puy Des Gouttes tephra has slumped off due to dike-triggered uplift.
Lemptégy 1 and Puy Des Gouttes
There is no sharp contact between the Lemptégy 1 agglomerates and Puy des Gouttes deposits, and Puy des Gouttes lapilli are found within the upper layers of the Lemptégy 1 deposit. The Puy des Gouttes scoria is distinguished only by its constant lapilli grain size, and it becomes easier to identify as it increases in relative abundance toward the top of the Lemptégy 1 agglomerate. The Puy des Gouttes tephra deposit is well sorted and layered.
The two Lemptégy 1 pahoehoe lava flows are found within Puy de Gouttes scoria. Thus, the two eruptions were coeval (De Goër de Hervé et al., 1999; Boivin et al., 2009a; Figs. 2 and 4A). The simultaneous activity is also confirmed by the red color of the Puy de Gouttes deposits right above the Lemptégy 1 agglomerates, which were hot enough to heat the freshly deposited Puy de Gouttes tephra. A few centimeters above this contact, the Puy de Gouttes pyroclasts become progressively blacker due to decreasing influence of the heat from Lemptégy 1 deposits. Lemptégy 1 can thus be seen as a satellite system of the larger Puy de Gouttes cone. The history of Lemptégy 1 is schematized in Figure 6.
Lemptégy 2 is a basaltic trachyandesitic to trachyandesitic cone (Table 1; De Goër de Hervé et al., 1999; Boivin et al., 2009a) that grew on the west side of Lemptégy 1 (Figs. 1C and 7A; Table 1). Even though Lemptégy 2 was thought to have erupted shortly after Lemptégy 1, and radiometric dates for these two volcanic centers are within error, an intermittent centimeter-thick weathered layer between the two formations indicates a gap in time. The climate then was subglacial and arid, so soil may have taken decades or even centuries to develop.
Lemptégy 2 deposits unconformably overlie Lemptégy 1, Puy des Gouttes deposits, and/or the weathered layer. A full record of Lemptégy 2 stratigraphy is accessible due to several quarry benches (Fig. 7B). Although some units are not continuous or are locally thinner than elsewhere due to Lemptégy 1 relief and wind directions, a correlation between units, and thus eruptive phases, is possible throughout the quarry (Fig. 8).
Lemptégy 2 deposits are mostly composed of scoriaceous lapilli and bombs of various shapes (fusiform, cowpat, bread crust) and variable vesicularity. The layers are generally well sorted (1 < σφ < 2), with little variation in chemical composition, although there is a very slight decrease of SiO2 content from lower to upper units (Fig. 7A; Wallecan, 2011). Reverse and normal grading are both observed, especially within the upper part of the deposit. Lemptégy 2 deposits consist of seven main units (Fig. 7B). The main units are composed of lapilli scoria and bombs. Some zones contain unbroken bombs, while other zones within the same unit have dominantly broken bombs.
The two thin key layers, units 2 and 4, are mainly made of <1 cm, highly vesicular pumice-like lapilli. The lowest, unit 2, is yellowish and composed of fragile, highly vesiculated, and low-bulk-density pumice (∼77% vesicularity; bulk density of 0.62 g cm–3; data from Wallecan, 2011), while the base of the overlying unit 3 is made of angular, denser black scoriae (46% vesicularity; bulk density of 1.45 g cm–3; data from Wallecan, 2011). The thickness of unit 2 is variable but does not exceed 10 cm. Unit 4 is also marked by a lighter color, and it is similar to unit 2 in terms of density and vesicularity. The transitions between these two key layers and the bounding units are not sharp, and trachybasaltic pumiceous lapilli are found in lesser amounts in the coarser surrounding layers, while a few bombs are present within the pumiceous lapilli layers. Units 2 and 4 are only present on the west side of the quarry.
The other key layer, unit 6, is higher up in the stratigraphic sequence of Lemptégy 2, and it is characterized by massive, dense bombs supported in a lapilli matrix, with some lateral variation in vesicle content of the lapilli. Small groupings of very large cowpat bombs are found locally. The lapilli are scoriaceous in the west side of the quarry, and they become angular and massive with very few vesicles toward the east. This unit can be found across most of the quarry.
Units 1, 3, and 5, which form the bulk of Lemptégy 2, are made of red and/or black lapilli and bombs of various sizes, shapes, and vesicle contents. Unit 1 contains several meter-wide cowpat bombs. Unit 7 is mostly composed of homogeneous material, i.e., well sorted (sorting coefficient varying from 1.33 to 1.44; Cas and Wright, 1987), and bombs are scarce. However, this unit is also marked by discontinuous pumiceous lapilli layers on the west side that become continuous toward the east. The pumiceous lapilli layers are marked by a white line on Figure 8. Near to the Lemptégy 1 edifice, and the upper 10 m of the Lemptégy 2 deposits, lenses of angular blocks and lapilli with coarse reverse grading are common, indicating local avalanching as the cone grew. Such avalanche layers dominate the breached area on the south side.
Xenoliths are found at several locations, either as free clasts or within bombs as angular or fluidal and partially melted xenoliths. The xenoliths are white and foamy and contain partially melted mica, amphibole, and glass. Their mineralogy is similar to the surrounding basement observed a few kilometers away from Lemptégy and on top of which the Chaîne des Puys is built. However, their alteration and melting preclude easy assignment to a precise original rock type. The xenoliths were thus probably incorporated in the upper crust and partially remelted during magma ascent. In Lemptégy 2, they are most abundant and largest (up to 30 cm across) in the upper units, especially in unit 6. Rare xenoliths are also found within the Puy des Gouttes deposit, and they are abundant in Lemptégy 1. We note that these xenoliths, being strongly melted, are quite different from the fresh ones reported by Valentine (2012) in the San Francisco volcanic field. On the other hand, partially melted and foamy basement xenoliths can be observed in other scoria cones, for example, at Parícutin (e.g., Rowe et al., 2011), at Mount Goma (Denaeyer, 1975), and at the Beaunit Maar to the north of Lemptégy (Jannot et al., 2005; Hardiagon et al., 2011).
Red-black color variations within Lemptégy 2 units are common, and we observed red-dominant and black-dominant patches. The red zones are adjacent to the larger intrusions and are likely to be related to heat-flux variations. We noted that red bombs could be found in black scoria and black bombs in red scoria.
Some units are locally deformed due to later intrusion emplacement. The deformation took place as folding of layers above intrusions, or, more rarely, as local faulting close to an intrusion boundary. To the south, there is a well-exposed major discontinuity in the quarry cliff. The layers are truncated, and the eastern side is composed of blocks and a small lava flow. This area is a cross section of the collapse breach that can be seen in the old topography of Lemptégy (Fig. 1C).
The Lemptégy 2 deposits indicate normal Strombolian venting, with explosions producing vesicular bombs and scoria. The deposits record mostly ballistic and jet emplacement (Riedel et al., 2003), with some tephra fall recorded in the finer lapilli fraction. Occasional large cowpat bombs may indicate that a small lava lake, or conduit filled by lava to the crater level, resided for short periods. There is a lack of rolled or composite bombs, which indicates little vent recycling, or that recycled bombs did not generally exit the crater. The two layers containing higher concentrations of more vesicular lapilli (units 2 and 4) may indicate short periods of increased fragmentation, possibly with more sustained violent Strombolian activity. We tentatively correlate these with cone-breaching episodes similar to the one suggested by Németh et al. (2011a) for the Los Morados cone. The dense lapilli in units 6 and 7, with the prismatic blocks found at the very top of the Lemptégy sequence, indicate that the vent was becoming choked by material and that milling of vent pyroclasts was occurring. The dense lapilli, associated with xenoliths in unit 6, may also indicate a minor phreatomagmatic quenching of some fragments at this stage. These late-stage deposits correlate with the presently exposed vent textures and indicate an increasingly Vulcanian aspect for the latter part of the eruption.
Lemptégy 2 Spatter Cones, Lava Flows, and Final Conduit
Quarrying of the volcano revealed information about the early composite activity of Lemptégy 2. An initial major lava flow that spread west and southwest occurred early in Lemptégy 2 history, as its base lies on just a few centimeters of scoria above Lemptégy 1 deposits (phase 1, Fig. 9). This main lava flow is the one on which the Vulcania Geoscience Park is now located. Several spatter-fed clastogenic flows were then emitted from 3 to 4 spatter cones, which are now excavated depressions a few meters deep (maximum 4 m) and ∼10 m in diameter. Their rims are usually made of massive, black, and glassy layers containing rare elongated vesicles, minor scoriae, and fiamme-like structures that may be flattened spatter. These layers are sometimes covered by a rough, elongated, vesicle-rich coating. Several small units show flow directions and dips toward the small vents (backward flows in Fig. 10). The ensemble is thus interpreted as spatter-fed flows (Johnston et al., 1997). Below the rims, scoriae are nonwelded or cemented by fine ash. Nonwelded scoriae, probably deposited later in Lemptégy 2 history, top the spatter cone rims.
The small spatter cones are bounded by parallel massive vertical walls mainly composed of welded products. These vertical walls are interpreted as eruptive fissures that propagated laterally from the spatter cones (eruptive fissure I, phase 2, Figs. 9 and 10). Friction structures along the inner walls are preserved (striae and tension cracks) and are interpreted as evidence of volcanic product flow-back related to a decrease in activity at the vent. Although the main activity was primarily concentrated around one spot (indicated by “1” in phase 2, Fig. 9), a small shift in the eruption focus occurred during the volcano’s growth, as shown by the formation of another eruptive fissure (eruptive fissure II) at the north of previous active spot “1.” Fissure II gave birth to a channelized lava flow that runs toward the northwest (indicated by “2” in phase 2, Fig. 9).
The layers seen in the outer part of the south and east quarry walls are tilted, faulted, and fractured. This deformation was directly linked to the underlying exposed cryptodomes. These deformed layers are further covered by draped scoria layers, indicating that the intrusive activity occurred during the growth of the cone. The cryptodomes are seen feeding small lava flows (phase 3, Fig. 9): three small lava flows originate from the southern cryptodome, and one from the northern one. The lower flows are highly fractured, indicating that the dome was growing during the lava effusion.
Other small lava flows are seen within the southern breached area. These were progressively uncovered during extraction between 2004 and 2007. They were sourced from a spatter vent within the breached zone. This is now destroyed, although the feeder dike is still visible, and a remnant of a lava tube extending from the vent is preserved in the Lemptégy Exhibit hall.
To the north of the early emplaced spatter cones, we observed a mixture of angular breccia, bread-crust bombs, prismatically jointed blocks, and rounded abraded blocks <1 m and larger (2–10 m) wide massive blocks of glassy rock (Fig. 11). The breccias and blocks are strongly baked and striated, and clasts are mostly rounded. Many of the blocks and layers are dipping inward. This formation is ∼20 m thick, and we interpret it as the final conduit, where the activity was concentrated toward the end of the eruption (phase 4, Fig. 9). The larger blocks within the conduit have sharp, spiny, or pahoehoe-like sides and are prismatically jointed. They are probably blocks or intrusions of viscous, degassed magma that were fractured during late-stage explosions and then were passively deformed. The angular breccias are broken up versions of these. The very uppermost layers of Lemptégy 2 contain similar rounded, abraded scoria and angular bombs, as well as the more characteristic Lemptégy scoria seen in lower units. This indicates that Vulcanian explosions dominated the terminal activity. Possibly, this was combined with or preceded by phreatomagmatic phases, as we observed xenoliths along with massive, dense bombs below the top unit of Lemptégy 2 (unit 6).
Lemptégy 2 Intrusions
The Lemptégy 2 shallow plumbing system is less clustered and reveals more structural and morphological features than that of Lemptégy 1. The plumbing system of Lemptégy 2 is composed of bulges, cryptodomes, and dikes. A summary description of intrusive morphologies can be found in Petronis et al. (2013). We observed a continuum of widths, from thin dikes of 40 cm to 5 m to swollen bulges and cryptodomes (near-surface bulges) that can reach up to ∼10 m in diameter (Fig. 12).
Macro- and microtectonic structures are abundant along intrusion walls. The quarrying has cut many intrusions, revealing their interiors, and these cross sections show the inner textures and structures, including tension gashes, shear zones, vesicles, and scoria incorporation (Fig. 13). All those structures are indicators of magma flow direction and emplacement mode. The main flow direction deduced from the structures is inward and dipping slightly upward (Delcamp et al., 2007), which was confirmed and complemented by AMS studies (Petronis et al., 2013). Dike sections show that the thinner dikes usually have shearing along one of the margins. The thicker dikes usually have shearing structures on opposite walls. Opposing shear zones converge toward the dike center (Figs. 12A, 12B, and 12C).
The bulges and cryptodomes show strongly sheared and brecciated zones and are associated with deformation of the overlying layers (Figs. 12B, 12D, and 12E). A few lavas extruded out of these superficial bulges (see Fig. 9 and “Lemptégy 2 Spatter Cones, Lava Flows and Final Conduit” section). The southern cryptodome in Figure 12E is responsible for at least one partial collapse, since a clear discontinuity is associated with the deformed overlying layers.
The shallow plumbing system of Lemptégy 2 was strongly influenced by Lemptégy 1, which acted as a buttress during the intrusion of Lemptégy 2 (Fig. 14A). The concentration of bulges is thus higher toward the east than toward the west, where magma could freely travel as dikes through the cone, in contrast to the east, where magma propagation was limited due to Lemptégy 1. Neither Lemptégy 2 dikes nor bulges have been observed in the area of Lemptégy 1. The structures of one dike in the south reveal magma flow from the intrusive center southward. This dike is linked along strike to the wall of the southern breach (see Fig. 9). The breach was exploited during quarrying, providing easy access to the cone. The breach can still be determined by the geometry of the dikes around the main central intrusive system, which formed a half-cup–like shape open toward the south (Fig. 14B). The collapse scar is also seen clearly in the quarry wall deposits.
Birth and Growth of Lemptégy
Eruptive activity began at Lemptégy 1 with the construction of a 30-m-high coarse bomb and scoria sequence at the base of the Puy de Gouttes, which was erupting simultaneously. Three main dikes were emplaced into Lemptégy 1, which converged toward the northeast, toward the Puy de Gouttes (Fig. 15). At least one of these dikes (dike 4) reached the surface, passing through the coarse bomb sequence, to feed proximal pahoehoe lava flows. The Lemptégy 1 eruptive activity stopped before the Puy des Gouttes volcano eruption ended, but intrusive activity continued, deforming erupted products, including those of the latter stages of the Puy de Gouttes eruption. The pahoehoe lavas from dike 4 were erupted after the bulging episode, as they are not deformed.
A gap long enough to form a few centimeters of poorly developed soil occurred before the renewal of eruptive activity. This new activity was focused toward the west side of Lemptégy 1. As indicated by the lack of Lemptégy 2 intrusions in Lemptégy 1, the latter acted as a buttress that partially controlled the morphology and evolution of Lemptégy 2’s shallow intrusive and extrusive systems.
The stratigraphy of Lemptégy 2 was defined using the western part of the quarry, and correlation between layers across the quarry has allowed the eruptive phases to be reconstructed as shown schematically in Figure 9. The sequence established for Lemptégy 2 is, however, incomplete, and it is condensed in the eastern part of the quarry due to the presence of Lemptégy 1. While Lemptégy 2 was growing, the relief of Lemptégy 1 acted as a barrier that impeded the eastward spread of the erupted products. We cannot exclude the influence of the wind, which could have also played a role in the dispersion of the fallout. The cone deposits, which are predominantly proximal ballistic clasts, had to surmount Lemptégy 1 in order to spread north and east. This is why only units 5, 6, and 7 could be traced across the quarry and also why they are condensed to the NE and E, away from the main Lemptégy 2 center.
The eruption of Lemptégy 2 began with a lava flowing into a paleovalley to the south and east, at the base of the Puy de Gouttes and Lemptégy 1. A main central fissure (fissure I) developed, and activity continued with Strombolian explosions from several vents that fed small lava flows. Activity subsequently concentrated into one central vent. As the cone grew, intrusions extended outward, and cryptodomes formed in the lower flanks. These breached the cone and fed small lava flows.
Layers of cowpat bombs, found mainly in units 1 and 6, attest to periods of probable small lava lake activity and associated short lava fountain events. The two layers rich in pumiceous lapilli (units 2 and 4) indicate an increase in explosivity. Because there is no sharp transition between pumiceous lapilli and surrounding units, the changes in explosivity are inferred to have been progressive. We suggest that this more intensive period corresponds to a phase of more lapilli- and ash-charged plumes, possibly a short sustained venting, or violent Strombolian event. This event may have been related to decompression linked to the progressive development of a landslide to the south, as suggested for the Los Morados cone, Argentina (Néméth et al., 2011a).
The latter stages of Lemptégy 2 became more explosive, as seen in unit 6, which has some very large bombs, and a dense angular fraction, which may indicate phreatomagmatic interaction. This is also the layer in which the largest granitic xenoliths are found, suggesting a link between these and the change in eruptive style. The unresolved question is whether the fragments were incorporated because of the increased explosivity, or if they caused it by adding volatiles to the magma during melting and ascension, as suggested for other volcanic systems (e.g., Hardiagon et al., 2011; Chadwick et al., 2013), or both.
The final Lemptégy 2 stage was focused on the main conduit, which had become choked with material, and a plug of magma formed. This was the most violent episode, as attested by the final upper layers of finer black ash, lapilli, and blocky bombs observed in the upper layers of Lemptégy 2. The breccias within the conduit and the uppermost layers of Lemptégy 2 suggest Vulcanian activity occurred during this last stage of Lemptégy growth.
During the cone growth, dikes were intruded into the southern flank, and it was breached. A new vent was established within the scar, and a small spatter cone and lava flow formed. This was later covered by clasts avalanching from the main vent, as the scar became partially infilled. The evolution of the cone’s shape changed the local stress field, and the newly intruding dikes were emplaced parallel to the collapse scar, i.e., perpendicular to σ3, as half-cup dikes (Fig. 14B). As the eruptive activity continued, other planar dikes were emplaced following a generally NNE-SSW direction. The final phase of activity was the development of a main central conduit, which buried the previous vents (Fig. 11). This conduit became progressively blocked by recycled products and degassed magma. The eruptive activity became sporadic, showing Vulcanian behavior, with discrete debris-charged explosions.
Most of the magma flow directions inferred from the microstructures along the intrusion walls and within the dikes, i.e., tension gashes and striations, indicate flow toward the main eruptive center. This is contrary to some other studies that have suggested outward flow alone (e.g., Hintz and Valentine, 2012). At Lemptégy, the inward flow direction probably reflects the primary flow rather than a late backward flow direction due to magma deflection, because structures showing inward flow, such as the shear zones, affect entire dikes, and some of these dikes formed early on. During the late stage of dike emplacement, the outer dike walls progressively cooled, and the deformation induced by a decreasingly pressurized magma would probably not have deformed the solid walls, affecting instead only the inner parts.
The flow toward the main eruptive center might indicate the presence of a deeper, wider conduit at depth below Lemptégy that was feeding magma toward the central conduit and focusing the latter eruption phases. Similar feeding systems have been described, e.g., Keating et al. (2008). The AMS data presented by Petronis et al. (2013) clearly suggest that this was the case.
As Lemptégy volcano grew, its load increased; at this stage, the volcano topography and its own local stress field would have begun to dominate the internal volcano shallow plumbing, as suggested for monogenetic cones in general by Valentine and Gregg (2008) and Valentine (2012). The crater and vent may also have been periodically choked with material, and degassed magma may have accumulated in the deeper conduit. Consequently, this magma had higher viscosity and low overpressure, and more work was required for it to propagate as a dike. Because of this, several dikes inflated to form thick dikes, with well-developed brittle-ductile structures and intense shearing. Some thick dikes received enough magma to approach the surface, where they developed into cryptodomes, which bulged the surface and occasionally erupted small lava flows.
Implications for the Chaîne des Puys and Relationship with Tectonics
The endogenous and exogenous growth of Lemptégy volcano was far from being straightforward and simple. Although it was a small system, barely 80 m high, this scoria cone showed multiple stages of evolution. The presence of a slightly weathered horizon between the two Lemptégy volcanoes indicates that a pause in activity of at least a few years occurred. The lapilli unit of the Puy de Gouttes between the Lemptégy 1 and 2 deposits also suggests concurrent activity, possibly indicating a deep connection between the three volcanic centers, similar to the Yucca Mountain range (Connor et al., 2000). This is also supported by the similar morphologies of the other volcanoes further along the alignment to the northeast (Puy de Jumes, Puy de Coquille; Fig. 1). The alignment of the Lemptégy-Gouttes-Chopine-Jumes-Coquille edifices might reflect one overall volcanic system with a common crustal magmatic reservoir feeding multiple monogenetic eruptions over a long time period (Fig. 15). The extended age range from Lemptégy 1 to the 8000-yr-old Chopine eruption could mean that the magmatic reservoir underlying this volcanic alignment was active over at least 20,000 yr. Confirmation of this hypothesis would require detailed petrological, geochemical, and isotopic studies. The longevity of the magmatic system has important hazard implications, as it extends the possible lifetime of each volcanic alignment in the Chaîne des Puys. Many of these have had eruptions within the last 20,000 yr and could thus be considered as potentially active magmatic reservoirs. Such a conclusion is in agreement with recent work on the trachytic rocks of the Chaîne des Puys, which suggests long-lived intrusive reservoirs (Martel et al., 2013).
Such long-lived reservoirs feeding monogenetic eruptions have also been inferred in other tectonic settings, such as in Central America and Mexico. For example, Cerro Negro (Nicaragua) is also part of a group of cinder cone alignments (van Wyk de Vries, 1993; McKnight and Williams, 1997; van Wyk de Vries et al., 2007). Such cone fields host magmatic systems that are long-lived, active over thousands of years, and evolve from basic to acid, similar to the Chaîne des Puys. Both in Central America and Mexico, there is active extensional or transtensional faulting, and often a close relationship exists between the volcanoes and the structures (Connor et al., 2000; Valentine and Krogh, 2006; Valentine and Perry, 2007; de la Cruz Reyna and Yokoyama, 2011).
A link between edifice alignment (with a main N-S trend and secondary NE-SW trend) and the local structure has been also highlighted in Lemptégy and elsewhere in the Chaîne des Puys (van Wyk de Vries et al., 2012). The dike orientations and the related flow trend at Lemptégy may reflect a local structural heritage, with NE-trending Hercynian dikes and shear zones (van Wyk de Vries et al., 2012) creating weak zones. In addition, the N-trending dikes reflect the general alignment of the Chaîne des Puys (Fig. 1). Thus, both of the two main regional trends appear in Lemptégy. There is a parallelism among the NE alignment of feeder dikes, the alignment of Lemptégy, the Puy de Gouttes and the volcanoes to the north, an area of basement shear zones to the northeast of the volcanoes, and eventually the Aigueperse fault in the Limagne graben. This suggests a possible Pleistocene tectonic control on the activity in the Chaîne des Puys. Thus, it may be that the monogenetic nature of the eruptions is related to the extensional tectonics and preexisting crustal structures, as suggested by van Wyk de Vries (1993) and van Wyk de Vries et al. (2007). The coincidence of extensional tectonics with a weak fracture zones would favor dispersed volcanism. Such a link was observed at the 1999 Cerro Negro eruption, which was accompanied by local rifting (La Femina et al., 2004).
Implications for Monogenetic Volcanoes: Plumbing Systems (Table 2)
Studies of scoria cone plumbing systems depend on the level to which the conduit is exposed. Most of the published papers concern levels that are deeper than 10 m below the original surface. At this depth, the plumbing system is seen to be reduced to a major conduit and a few feeder dikes (e.g., Keating et al., 2008; Valentine, 2012; Hintz and Valentine, 2012). Lemptégy quarry exposes intrusive and eruptive products from the final volcano surface to a level around 30 m below the pre-edifice ground level. This represents the missing link between intrusive plumbing system at shallow depths and that in the cone itself. Similarly, it explains the morphological differences between the larger and thicker dikes of exposed shallow plumbing systems observed at greater depth below the surface. The dikes at Lemptégy have a much greater range of thicknesses and are more irregular than previously described feeder dikes (Keating et al., 2008; Valentine, 2012; Hintz and Valentine, 2012).
Depending on the active stress field, the dikes in the upper crust usually follow existing faults (Connor and Conway, 2000; Valentine and Krogh, 2006) until they reach the near surface or the developing cone. At this shallow level, dike propagation depends on the local stress field induced by the topography (Hintz and Valentine, 2012). The intrusive system of Lemptégy 2 shows both types of dikes (Fig. 15): It shows the transition from dikes controlled by regional and crustal factors to those relating to the edifice structure and lithology. The feeder dikes show regional trends that guided the early orientation of eruptive fissures, and as they propagated up into the growing cone, they caused structures like small flank collapses and secondary vents. Other dikes were intruded to the west side of Lemptégy 2 from the conduit and formed a fan-shaped pattern of radial intrusions. As Lemptégy 2 intrusions are not found on the Lemptégy 1 (east side of Lemptégy 2 conduit), it is likely that they were controlled by buttressing and the edifice stress regime of Lemptégy 1.
Lemptégy 1 dikes and structural orientations roughly follow the regional trends; thus, the intrusions that formed the uplift seen in Figure 5 were also regionally controlled and follow the general trend of the Puy de Gouttes–Jumes and Coquille ridge.
Such breakouts at the bases of cones are common and have been observed at Cerro Negro (1967 and 1999, for example: Hill et al., 1998; La Femina et al., 2004), and at Paricutin, where the Salipichu vents are probably an equivalent to the Lemptégy 2 feeder dike vents. The Quitzocho ridge at Parícutin (Luhr and Simkin, 1993) could be an equivalent to the Lemptégy 1 uplift.
Implications for Monogenetic Volcanoes: Eruptive Dynamics
Key comparison points between Lemptégy and historical eruptions or other excavated cones are summarized in Table 3. Next, we discuss a few points, starting with the initial and final activity.
In many documented cases, the start of a monogenetic eruption begins with a phreatomagmatic phase, followed by Strombolian activity (Table 3; e.g., Vespermann and Schmincke, 2000; Schmincke, 2004; Rapprich et al., 2007; Kiyosugi et al., 2014). Lemptégy 1 does contain abundant partially melted crustal xenoliths, which have been associated with the early phreatomagmatic stages of some Chaîne des Puys eruptions (Hardiagon et al., 2011). However, the deposits show no evidence of phreatomagmatic quenching or fragmentation. This would imply either that the quarry did not cut deeply enough into the volcano so the phreatomagmatism phase is not revealed or Lemptégy 1 did not have such initial activity.
For the final stage of eruptions, a more explosive state has been suggested, for example, at Parícutin (Erlund et al., 2010), and Schmincke (2004) also described black, well-sorted lapilli as a final deposit of many scoria cones in the Eiffel volcanic field, but related them to subplinian activity. Lemptégy 2 exhibits a similar final phase to Parícutin, and conduit blocking with consequent Vulcanian activity may be a common feature of the final stage of such eruptions.
Violent Strombolian eruptions with sustained columns have been previously described from the deposits of monogenetic scoria cones such as Lathrop Well (USA; Valentine et al., 2007), Irao (Japan; Kiyosugi et al., 2014), and Parícutin (Luhr and Simkin, 1993). Similar activity is suggested by units 2 and 4 at Lemptégy volcano (Table 3).
The growth of Lemptégy is complicated, with numerous intrusive and extrusive phases, activity in many ephemeral vents, and shifts in eruptive style. Such diversity has been also observed during historical eruptions and interpreted from deposits in excavated or eroded cones such as Parícutin, Tolbachik, and Lathrop Wells (Valentine et al., 2007; Erlund et al., 2010; Gordeev et al., 2013). Another example is the shallow-level emplacement of cryptodomes that has been observed within the crater in Cerro Negro in 1995 (e.g., in Petronis et al., 2013). In Lemptégy, such shallow bulges have been observed in the western flank as well as within the final conduit. Such diversity of concurrent intrusive and extrusive activity in a basaltic volcano with a restricted range of composition can be attributed to the crystallization and degassing processes that change the magma viscosity as suggested for Lathrop Wells (USA; Valentine et al., 2007; Genareau et al., 2010). For both Lemptégy 1 and 2, we suggest that fresh gas-rich magma was fed in via feeder dikes, while magma was degassed in the conduit and then intruded outward from the center. When feeder dikes breached the surface, fluid magma could then escape directly, as in the Lemptégy 2 lava-flow–triggered breaching, or the Lemptégy 1 pahoehoe flows.
In many scoria cones, the crater and flank facies are separated from each other by semicircular faults or funnel-shaped surfaces (Schmincke, 2004). This distinction is not so clear in Lemptégy 1 and 2. In Lemptégy 2, the absence of well-defined wall versus crater facies might be due to the simultaneous activity of several spatter cones and cryptodomes. The final Vulcanian conduit is quite wide (50 m) and has a well-defined funnel shape, which is infilled with blocks and bombs. Possibly this conduit widened enough to destroy the previous crater margin structure. The absence of outer and inner crater distinctions at Lemptégy 1 may be due to the strong subsequent deformation that has obscured these features.
Although monogenetic cones share similar eruptive styles and activity, it is not yet possible to develop a general model that would be valid for all the edifices. Indeed, while some started with phreatomagmatic phases, other started with Strombolian/Hawaiian or effusive activity. Activities may be similar, but chronological order differs, and it appears that each monogenetic volcano has its own development phases, and to generalize a simple generic model would be unrealistic. We propose that the local environment (tectonic context, presence of crustal weak zones) plays an important initial role and that the volcano itself plays a major subsequent role in its own evolution: The eruptive dynamics and the intrusion and deformation phases interact with each other as the volcano grows (Fig. 9). For example, the growth of an intrusion will change the local stress field, which will condition the geometry and style of the next intrusive and eruptive phases. Deformation induced by intrusions can lead to small destabilization events that can influence or even disrupt the shallow plumbing system. Similarly, the morphology of the cone itself plays a role in the volcano’s evolution. Such complexities are reflected in the various morphologies of scoria cones and underline the fact that the variability in eruptive style strongly influences the final edifice shape and subsequent erosion (Martin and Németh, 2006; Németh et al., 2011b; Kereszturi et al., 2012; Kervyn et al., 2012; Kereszturi and Németh, 2012). The modification of stress following the Lemptégy flank collapse influenced the endogenous growth, but on other scoria cones, such events have been associated with a change in eruptive dynamics (Németh et al., 2011a).
The Lemptégy volcano has been excavated down to reveal the shallow feeder system of two scoria cones, Lemptégy 1 and 2. This provides a remarkable opportunity to observe the growth of two monogenetic cones from their surface down to their roots. The detailed analysis of the shallow plumbing system and associated volcanic deposits of Lemptégy leads to several key conclusions.
(1) Lemptégy, and many recent works, shows that the simple morphology of a monogenetic scoria cone can hide significant internal complexities, resulting from many interactions that take place during the edifice growth. Certainly, the early descriptions of Lemptégy as “insignificant and simple” (e.g., Scrope, 1858) show that the surface morphology gave little clue as to the variety of events hidden below.
(2) The shallow plumbing system at Lemptégy does not comprise a single, simple conduit that fed lava flows and emitted tephra and bombs, but it is composed of a complicated and evolving system of dikes, conduits, cryptodomes, bulges, and vents. The interplay between intrusive, extrusive, and instability events made this system a dynamic and continuously changing environment. Lemptégy volcano is monogenetic in that each eruption represents one batch of magma, but it has a history of multiple processes combined over an extended time period, and possibly taking place over decades. Thus, the name “monogenetic volcano” is correct in terms of single magma pulses but belies the highly variable origin of the landform.
(3) This time scale is also quite relevant in terms of hazard assessment, as monogenetic fields develop over a long time range (sometimes over several millions of years), and new edifices usually grow up where no cone or dome existed previously. This, with the added dimension that monogenetic volcanoes can be highly variable in their type of activity, reacting to local ground conditions and to their own growing edifice, means that hazards from a monogenetic volcano will also be complex and may change rapidly. This poses a significant challenge to monitoring, and for mitigation strategies. As monogenetic volcanoes, by their nature, do not provide a previous history in deposits, each case will provide a unique challenge, which can only be dealt with by characterizing the broad range of potential outcomes. Any response would need to be highly adaptable to the rapid changes.
(4) It is significant that the complexity at Lemptégy has only come to light through excavation. Importantly, this shows the value of quarrying to research and geological heritage in certain cases. The history of extraction at Lemptégy also shows that intelligent extraction can reap great benefits and can raise the value of the final excavation. As quarrying is ongoing in most monogenetic volcano fields worldwide, Lemptégy should serve as an example of both the benefits to be gained and of the way to achieve them.
We would like to warmly thank the Lemptégy staff for discussions and for giving us open access anytime to the quarry. We are also thankful to (many) volcalcanologists who came to discuss on the field: J. Sumner, T. Walter, O. Melnick, C. Seibe, S. Cronin, G. Valentine, S. Self, W.I. Rose, T. Thordarson, P. Grosse, D. Rothery, J. Calvero, S. Carn, M. Petronis, J. Lindline, Teddy Wolf, N. Riggs, M. Ort, A. Marquez, O. Galland, G. Ernst, M. Giardino, B. Ward, E. Calder, V. Troll, P. Byrne, E. Holohan, R. Herrera, and A. Foulks. Special thanks go to C. Olive Garcia and her team from the Conseil Géneral du Puy De Dôme. We also would like to thank Fran van Wyk de Vries for efficient proofreading of the manuscript. Finally, we would like to thank an anonymous reviewer, G. Valentine, and Editors F. Mazzarini and S. de Silva for the positive and constructive reviews provided.