Tectonics, hydrothermalism, and paleoclimate recorded by Quaternary travertines and their spatio-temporal distribution in the Albegna basin, central Italy: Insights on Tyrrhenian margin neotectonics

Hydrothermal settings are common in many tectonically active continental regions (Browne, 1978; Bibby et al., 1995; Rae et al., 2003; Newell et al., 2005; Crossey et al., 2006; Uysal et al., 2009; Cas et al., 2011; Mazzini et al., 2012; Baillieux et al., 2013; Karlstrom et al., 2013; Ricketts et al., 2014; Sella et al., 2014), where active faulting and fracturing provide viable pathways for circulation and mixing of endogenic and meteoric fluids, leading to diffuse mineralization and hydrothermal outflow (Curewitz and Karson, 1997; Cox et al., 2001; Rowland and Sibson, 2004; Billi et al., 2007; Gudmundsson, 2011; Bigi et al., 2014; Crossey et al., 2015; Vignaroli et al., 2015).

In hydrothermal settings where carbonate-enriched fluids circulate within calcareous reservoirs, thermogene travertine is the common CaCO₃ sinter precipitated from thermal springs (Pentecost and Viles, 1994; Pentecost, 1995). Thermogene travertine deposits have been documented to represent important markers for the mode and style of tectonic activity within hydrothermal settings (e.g., Altunel and Hancock, 1993; Hancock et al., 1999; Altunel and Karabacak, 2005; Uysal et al., 2007; Faccenna et al., 2008; Brogi and Capezzuoli, 2009; Brogi et al., 2010a; De Filippis and Billi, 2012; De Filippis et al., 2013b; Frery et al., 2015). Moreover, thermogene travertine has been used as a reliable indicator of paleoclimatic oscillations (Sturchio et al., 1994; Rihs et al., 2000; Soligo et al., 2002; Faccenna et al., 2008; Uysal et al., 2009; De Filippis et al., 2013a; Toker et al., 2015) and paleohydrological regimes (Crossey et al., 2006; Crossey and Karlstrom, 2012; Priewisch et al., 2014).

The region of Tuscany in central Italy (Fig. 1) is characterized by diffuse fossil and active hydrothermalism associated with highly productive geothermal areas of the Larderello-Travale and Mount Amiata fields (Batini et al., 2003). These geothermal fields are characterized by heat flux higher than 600 mW·m^–2 (Della Vedova et al., 2001) and collectively provide an annual electrical production of more than 5300 GWh (Bertani, 2005). Hydrothermalism in Tuscany is originated by high heat flow due to Miocene–Quaternary postorogenic thinning of the central Apennines chain (present-day crustal thickness is ∼22–24 km; Locardi and Nicolich, 1988; Billi et al., 2006) and associated emplacement of dominantly anatectic products of the late Miocene–Pleistocene Tuscan magmatic province (Innocenti et al., 1992; Marinelli et al., 1993; Serri et al., 1993; Carmignani et al., 1994; Jolivet et al., 1998). Active hydrothermal springs and gas manifestations, including active travertine-depositing springs (e.g., Minissale, 2004; Barazzuoli et al., 2013; Capezzuoli, 2013), are the main present-day features of ongoing hydrothermal activity, particularly in southern Tuscany (Figs. 1 and 2). Recent studies of travertine and hydrothermal ore deposits have documented structurally controlled fluid flow in the region (e.g., Bellani et al., 2004; Brogi and Fabbrini, 2009; Brogi et al., 2010a; Liotta et al., 2010; Rossetti et al., 2007, 2011; Rimondi et al., 2015; Croci et al., 2016; Berardi et al., 2016). In particular, Pliocene–Pleistocene faulting has played a primary role in the endogenic fluid circulation that fed and still feeds some geothermal fields and CaCO₃-rich springs in the Mount Amiata geothermal area (Brogi et al., 2012). This geological setting makes Tuscany an excellent area for studying the relationships between hydrothermalism and tectonics, and their spatio-temporal evolution, through thermogene travertine deposits.

This paper addresses the relationships among Quaternary hydrothermalism, tectonics, and paleoclimatic evolution in the Neogene–Quaternary Albegna basin of southern Tuscany (Fig. 2). Our multidisciplinary approach integrates structural investigations, geochronological analyses (²³⁰Th/²³⁴U, uranium-series disequilibrium), and stable isotope (δ¹³C and δ¹⁸O) systematics on selected carbonate structures (bedded and banded travertines and calcite-filled veins). The main objectives are (1) to reconstruct the spatio-temporal evolution of a hydrothermal system in a tectonically controlled area of geothermal interest, and (2) to discuss possible implications for the neotectonic regime along the Tyrrhenian margin in central Italy.

Southern Tuscany is part of the hinterland domain (on the Tyrrhenian Sea) of the Paleogene–Quaternary Northern Apennines chain (Fig. 1), a fold-and-thrust belt resulting from Mesozoic–Cenozoic tectonic convergence between the European and African plates (e.g., Dewey et al., 1989; Boccaletti et al., 1990; Wortel and Spakman, 2000). Orogenic construction of the Apennines occurred simultaneously with the westward subduction of oceanic lithosphere and progressive involvement of the Adriatic (African affinity) continental margin (Royden et al., 1987; Doglioni, 1991; Faccenna et al., 2004; Rosenbaum and Lister, 2004). The growth of the Apennines involved a general eastward migration of thrust fronts and foredeep basins in a classical piggyback sequence toward the foreland (e.g., Patacca et al., 1990; Cipollari and Cosentino, 1995; Massoli et al., 2006). Since Miocene time, a postorogenic extensional regime has occurred in the hinterland (Tyrrhenian side) domain of the Apennines, producing crustal-scale extensional fault systems, which have dissected the thick orogenic pile (e.g., Malinverno and Ryan, 1986; Jolivet et al., 1998; Cavinato and DeCelles, 1999).

In southern Tuscany (Fig. 2A), Miocene-to-Quaternary postorogenic sedimentary sequences (Martini and Sagri, 1993; Liotta, 1994; Pascucci et al., 2006; Brogi and Liotta, 2008; Brogi, 2011; Brogi et al., 2013, 2014) unconformably overlie a tectonic nappe stack composed of, from top to bottom (Carmignani et al., 2013, and references therein): (1) oceanic-derived units of the Ligurian domain, consisting of marly-arenaceous flysch-type and discontinuous ophiolitic sequences (Lower Cretaceous to Upper Eocene in age), and (2) continental-derived units of the Tuscan domain, including a nonmetamorphic succession (Mesozoic carbonate and Cenozoic marly and siliciclastic sedimentary sequences of the Tuscan nappe) and underlying metamorphic units of the Tuscan metamorphic domain. These units are presently exposed in NNW-SSE–trending (as in the Mount Cetona area) and N-S–trending (as in the Mount Labbro area) elongate structural highs bounded by extensional and transtensional faults (e.g., Brogi, 2004a; Bonciani et al., 2005; Brogi and Fabbrini, 2009; Carmignani et al., 2013; Fig. 2B).

There are two opposing interpretations of the Neogene–Quaternary tectonics of southern Tuscany. The most common interpretation favors a postorogenic extensional regime acting at the rear (west) of the eastward-migrating compressional front (e.g., Carmignani et al., 1994; Keller et al., 1994; Lavecchia et al., 1994; Faccenna et al., 1997; Barchi et al., 1998; Jolivet et al., 1998; Martini et al., 2001; Collettini et al., 2006). This regime has led to the formation of orogen-parallel extensional basins, where NW-striking basin-boundary faults have been dissected by transverse transfer zones, the latter accommodating differential rates and amounts of extension between adjacent extensional compartments (Liotta, 1991; Faccenna et al., 1994; Acocella and Funiciello, 2002, 2006; Liotta et al., 2015). Alternatively, a Miocene–Pliocene shortening regime has been described as having been active in Tuscany, causing basement duplexing and out-of-sequence thrusting (e.g., Boccaletti et al., 1997; Cerrina Feroni et al., 2006; Musumeci and Vaselli, 2012; Bonini et al., 2014). In this latter interpretation, extensional and strike-slip faulting would represent the most recent mode of deformation in Tuscany.

Late Miocene–Quaternary magmatism is localized along the Tyrrhenian margin (Fig. 1; e.g., Peccerillo, 2003; Conticelli et al., 2015) and characterized in Tuscany by acidic intrusive and volcanic products with associated high-temperature metamorphism (e.g., Barberi et al., 1971; Innocenti et al., 1992; Serri et al., 1993; Acocella and Rossetti, 2002; Rocchi et al., 2002; Dini et al., 2005; Rossetti et al., 2007, 2008; Farina et al., 2010; Cifelli et al., 2012). The Tuscan magmatic province hosts fossil and active hydrothermal systems. Endogenic fluid circulation within hydrothermal systems has been dominantly channelized by possibly active extensional faults (e.g., Barberi et al., 1994; Buonasorte et al., 1988; Chiarabba et al., 1995; Gianelli et al., 1997; Batini et al., 2003; Bellani et al., 2004; Annunziatellis et al., 2008; Brogi, 2008; Brogi et al., 2010b, 2015; Liotta et al., 2010; Rossetti et al., 2008, 2011).

The Mesozoic carbonate units of the Tuscan nappe have exerted a pivotal role in the functioning of the entire geothermal-hydrothermal setting of Tuscany. At surface and shallow levels, these rocks provide the recharge areas where meteoric waters infiltrate to depth thanks to the well-developed fault-fracture permeability network. At depth, carbonate rocks constitute the reservoirs where meteoric and endogenic fluids circulate and mix before ascending to feed surface thermal springs and CO₂ emission centers (e.g., Batini et al., 2003; Minissale, 2004). Active and fossil travertine deposits occur with variable size and shape in proximity to exposed Mesozoic carbonates and at the peripheries of main volcanic centers (Mount Amiata and Vulsini Mountains in Fig. 2A). As explained already, these travertine deposits originated from the long-term interactions among faults, fractures, and hydrothermal fluids during the Quaternary (Brogi, 2004b; Brogi et al., 2012).

The Albegna basin (Fig. 2A) is delimited to the north by the Mount Amiata volcanic district (300–190 ka; Cadoux and Pinti, 2009; Laurenzi et al., 2015; Marroni et al., 2015) and to the southeast by the Vulsini Mountains volcanic district (590–127 ka; Nappi et al., 1995). The Albegna basin is filled by marine and transitional sediments, which are late Miocene to Quaternary in age. These deposits consist of marine clays, regressive sands, gravels, and conglomerates covered by eolian sands and fluvial clays (Zanchi and Tozzi, 1987; Bettelli et al., 1990; Bonazzi et al., 1992; Bossio et al., 1993, 2003). Steeply-dipping, N-S–, E-W–, and NE-SW–striking tectonic structures, mainly consisting of extensional and oblique to strike-slip faults, dissect the basin-filling deposits (Fig. 3A; Zanchi and Tozzi, 1987; Brogi, 2004a; Bellani et al., 2004; Brogi and Fabbrini, 2009). Both extensional and strike-slip fault systems disarticulated the substratum, exposing the Jurassic units to the north of the study area (Fig. 3B; e.g., Bettelli et al., 1990; Brogi, 2004a; Bonciani et al., 2005; Carmignani et al., 2013; Guastaldi et al., 2014).

Both regional uplift (related both to regional tectonics and to the Mount Amiata volcanic bulging) and eustatic fluctuations contributed to the morphological shaping of the Albegna basin during the Pleistocene–Holocene period (Piccini et al., 2015). In fact, marine Pliocene sediments presently occur up to ∼600 m above sea level (a.s.l.), and multiple Quaternary alluvial terraces occur at different elevations between 50 and 300 m a.s.l. The present landscape of low rolling hills is dominated by the alternation of morphological depressions (mostly filled by Quaternary deposits) and positive morphotypes, both being affected by well-pronounced escarpments and canyon incisions.

As demonstrated by the geochemistry of numerous thermal springs, the hydrogeological setting of the Albegna basin and surrounding areas is mainly conditioned by the deep aquifer occurring within the Mesozoic carbonate units of the Tuscan nappe (e.g., Baldi et al., 1973; Duchi et al., 1992; Chiodini et al., 1995; Minissale, 2004). This aquifer has experienced significant vertical oscillations during the Quaternary, as also shown by the occurrence of speleogenic markers and landscape changes (Piccini et al., 2015). Presently, the general southward drainage is toward the Saturnia thermal springs (Fig. 2A), where gas emissions and travertine deposition are still active (Minissale, 2004).

In the northern part of the Albegna basin, a series of travertine deposits lies unconformably on top of the Neogene sediments (Zanchi and Tozzi, 1987; Bosi et al., 1996; Barilaro et al., 2012). Some of these travertine deposits are affected by faults with associated damage zones and joint networks. These deformations have been considered as resulting from either Pliocene–Pleistocene extensional tectonics (Zanchi and Tozzi, 1987), or alternating contractional and extensional tectonic phases active in southern Tuscany during the late Pleistocene (Martelli et al., 1989). Based on morphological and stratigraphic characteristics of the Albegna travertine deposits, Bosi et al. (1996) proposed discrete phases of travertine deposition over a long time interval between Messinian and Holocene times. So far, the only available radiometric age (218 +39/–27 ka) for the Albegna travertine deposits was determined on a travertine sample collected in a quarry located immediately to south of the village of Saturnia (Taddeucci and Voltaggio, 1987). Additional radiometric dating is thus required to constrain (1) the ages of these travertine deposits and (2) their possible relationships with the hydrothermal, tectonic, and climatic conditions within the Albegna basin and the greater Tuscan region.

Structural investigations were carried out along a N-S–trending transect in the northern part of the Albegna basin, where series of travertine deposits are exposed. From north to south, these deposits are named: Semproniano Ridge, Poggio Semproniano, Poggio i Piani, Pian di Palma, Poggio Pancotta, Saturnia village, Bagni di Saturnia, and I Pianetti (Fig. 4A; Table 1). Field observations were focused on the recognition of: (1) the different travertine morphotypes (plateau vs. fissure ridge travertines), (2) the different styles of travertine deposition and precipitation (bedded vs. banded travertines), (3) the geometric relationships between the travertine deposits and the surrounding units, and (4) the structural features (fault and fracture systems) postdating travertine deposition. Fault and fracture systems were studied in terms of their geometry (attitude, spacing, aperture, persistence, and crosscutting relationships) and kinematics. Results of our field investigations (Table 2) are synthesized in the geological-structural map of Figure 3A.

Fissure ridges are defined as elongate mound-shaped travertine deposits, straight or curved in map view, with a main crestal fissure, and length spanning from a few meters to several hundreds of meters (e.g., Hancock et al., 1999; Altunel and Karabacak, 2005; Brogi and Capezzuoli, 2009; De Filippis and Billi, 2012; De Filippis et al., 2013b). Travertine plateaus are defined as travertine deposits characterized by centimeter-thick, subhorizontal bedding. Travertine plateaus are tabular bodies roughly equidimensional in map view, often resulting in topographic highs.

Bedded travertines are the primary travertine strata formed during open-air CaCO₃ precipitation from saturated H₂O solutions. Banded travertine are the CaCO₃ precipitates filling fractures that cut through the bedded travertines (and also other host rocks) or develop in a sill-like fashion along the travertine beds themselves. These structures are usually filled by sparitic and variably colored bands of CaCO₃ precipitated in a non-open-air (intralithic) environment (e.g., Uysal et al., 2007; De Filippis et al., 2013a). We use the term veins only for fractures filled by nonbanded sparry calcite with maximum thicknesses of a few centimeters.

Bedded and banded travertines, veins, and speleothem-like concretions were systematically sampled across the study area for geochronological and isotopic analyses. Samples were dated using the ²³⁰Th/²³⁴U method (Ivanovich and Harmon, 1992), which is based on the fractionation of the parent isotopes ²³⁸U and ²³⁴U from their long-lived daughter ²³⁰Th. This technique assumes that thorium is not included in the crystal lattice of the carbonate at the time of deposition, being easily hydrolyzed and precipitated or adsorbed on the detrital fraction, whereas uranium is soluble in the surface and near-surface environments, coprecipitating with CaCO₃ upon exsolution of CO₂. A ²³⁰Th/²³⁴U ratio close to one indicates that ²³⁰Th and ²³⁴U have reached secular equilibrium and therefore give an age older than the ca. 350 ka limit of the ²³⁰Th/²³⁴U dating method. The ¹³C/¹²C (δ¹³C) and the ¹⁸O/¹⁶O (δ¹⁸O) ratios of samples were investigated to distinguish between thermogene and meteogene carbonates and to characterize the origin and properties of the parental fluids (Friedman, 1970; Manfra et al., 1974; Fouke et al., 2000; Pentecost, 2005; Kele et al., 2011). In thermogene travertines, CO₂ mainly derives from deep magmatic fluids and from their interaction with carbonate rocks. Conversely, CO₂ in meteogene travertines mainly derives from the atmosphere and from shallow deposits such as soils (Turi, 1986; Kele et al., 2003; Pentecost, 2005). These distinct CO₂ sources are reflected in the δ¹³C values, with thermogene travertines being characterized by δ¹³C values between −3‰ and +8‰ and meteogene travertines being characterized by average δ¹³C values of −8.48‰ (Pentecost, 2005). Eventually, from the δ¹⁸O values, we calculated the temperature of the travertine and vein parental fluids using the equation of Kele et al. (2015).

Fossil and active travertine deposits of the Albegna basin are aligned roughly N-S along an ∼18 km transect (Fig. 4A) and are exposed at different elevations (Fig. 4B). The northernmost travertine deposits are represented by the isolated fissure ridge forming the bedrock of the Semproniano village at ∼550 m a.s.l. and two thick plateaus (Poggio Semproniano and Poggio i Piani) forming positive morphological structures at ∼600–700 m a.s.l. Moving southward, an additional travertine deposit is exposed in a morphological depression in the Pian di Palma locality (∼230 m a.s.l.) on the northwest side of the Albegna River. On the southeast side, two travertine deposits form tabular positive morphological features in Poggio Pancotta (∼200 m a.s.l.) and Saturnia (250–290 m a.s.l.). The southernmost travertine deposit, which is exposed at I Pianetti (∼230 m a.s.l.), forms a roughly tabular feature filling a morphological depression. Active hydrothermalism and travertine deposition occur at ∼140 m a.s.l. at Bagni di Saturnia. In the study area, this is the travertine deposit with the lowest elevation.

Studied travertine deposits were analyzed in terms of their morphological characteristics to understand the travertine morphotypes and their associated internal fabrics. The following outline refers to the field sites of structural measurements shown in Figure 3A. The Semproniano travertine (site 1) consists of a NW-SE–striking fissure-ridge structure (Berardi et al., 2016) located 500 m to the north of the Poggio i Piani and Poggio Semproniano travertine plateaus (sites 2 and 4, respectively; Fig. 5A). The exposed ridge is ∼700 m long in the NW-SE direction and 400 m wide in the perpendicular direction. In the northwestern part, the ridge is in contact with Lower Cretaceous–Oligocene shales and marls of the Scaglia Toscana Formation (Tuscan nappe) and Pliocene–Quaternary marine clays and sandstones (Fig. 5B). In the southeastern part, the ridge is in contact with Upper Jurassic radiolarites and siliceous shales (Diaspri Formation of the Tuscan nappe; see also Gelmini et al., 1967). The structural relationships between the travertine deposit and the surrounding units are not clear due to the Quaternary sedimentary cover and anthropic backfill. The internal fabric of the travertine fissure ridge in the Semproniano village consists of a wide (50 m at least), vertical to subvertical banded travertine (Figs. 5C and 5D) extensively exposed below the Aldobrandeschi Fortress. The banded travertine consists of alternating centimeter-thick white bands (with calcite crystals growing perpendicular to the wall) and gray-colored finer-grained bands. These bands are generally parallel to one another and strike parallel to the fissure ridge (average strike N315°), dipping subvertically N70° or N210° (Fig. 5D). In places, intersecting bands create V-shaped geometries (Fig. 5C). Poorly preserved remnants of subhorizontal bedded travertine are exposed on the distal part of the southwestern flank of the fissure ridge. This bedded travertine consists of plane-parallel brown-colored centimeter-thick beds (Fig. 5E). Unlike the banded travertine, which is characterized by nonporous sparry calcium carbonate, the bedded travertine is characterized by calcite shrubs, laminations, and millimeter-to-centimeter–sized cavities of both syndepositional and postdepositional (karst) origin (Fig. 5F). The contact between the banded and bedded travertines is not clearly visible.

About 1 km to the northwest of Semproniano, an isolated deposit of travertine is exposed in the I Vignacci locality (site 14). The internal fabric of this deposit consists of banded travertine with NW-SE–striking bands (Fig. DR1A¹), dipping steeply (∼55°) to the N240°, at high angle with respect to the adjacent subhorizontal Pliocene–Quaternary sequence. The orientation of this banded travertine is very similar to the one observed in Semproniano (average strike: N145°). This evidence suggests a structural and geometric continuity between the I Vignacci and Semproniano village banded travertines.

At Poggio Semproniano and Poggio i Piani, the travertine plateaus are subhorizontal (Figs. 5A and 6A) and lie on top of the Pliocene–Quaternary marine sequence or on the Scaglia Toscana Formation. They consist of plane-parallel (Fig. 6B), centimeter-thick beds of white-colored lime-mudstone with heterogeneous porosity due to the presence of microbialites and millimeter-to-centimeter–sized karst-dissolution cavities (Fig. 6C). Bedding is generally horizontal (Fig. 6C), although locally complicated by fault-related tilting (see following).

The Pian di Palma (site 5) and I Pianetti (site 10) deposits consist of subhorizontal travertine plateaus filling local depressions. The Pian di Palma plateau has a horseshoe-like shape of ∼6 km² in areal extent. An ∼20 m thickness of this travertine is exposed in an abandoned quarry at Pian di Palma (Fig. 6D). The quarry exposure is characterized by thick beds (average thickness between 20 and 50 cm) affected by numerous postdepositional features such as calcite veins, subhorizontal karst cavities, and vertical karst conduits. The stratigraphic-structural relationships between this travertine plateau and the adjacent units are not well exposed. The travertine is laterally in contact with (and probably lies on top of) the Pliocene–Quaternary marine sequence, the architecture of which is controlled by NW-SE–striking and N-S–striking extensional faults (Fig. 3A).

The travertine plateau at I Pianetti has a subcircular shape and a maximum thickness of ∼30 m, and it is visible in a large active quarry. The travertine is characterized by decimeter-thick, plane-parallel beds showing locally complex convoluted geometries.

At Saturnia (sites 7 and 8) and Bagni di Saturnia (site 9), the travertine deposits lie on top of the Pliocene–Quaternary marine sediments and mainly consist of bedded travertine boundstone with homogeneous primary porosity and millimeter-to-centimeter–sized karst-dissolution cavities. At the active travertine site of Bagni di Saturnia, different depositional facies are distinguished, including cascades, pools, and terraced slopes (Figs. 6E and 6F).

We studied the postdepositional structures affecting the travertine deposits in order to understand their geometry, spatial distribution, and tectonic and/or hydrothermal significance.

Fault sets with average orientations of N280°/75° and N75°/75° crosscut the northern limits of the exposures at Poggio Semproniano (site 3) and Poggio Pancotta (site 6), respectively (Fig. 7). At Poggio Pancotta, these fault sets are composed of damage zones (e.g., Caine et al., 1996) a few meters thick at the most (Fig. 7A). Fault planes dip steeply (> 50°) toward the NW or SE. Most slickenlines and abrasive striations on fault planes (Figs. 7B and 7E) are characterized by a pitch smaller than 10° or greater than 170° (see the stereographic projection in Fig. 7A), providing evidence for strike-slip–dominated kinematics. At Poggio Semproniano, an ∼3-m-thick damage zone (Figs. 7C and 7D) is characterized by 30-cm-spaced fault surfaces and NNW-SSE–striking synthetic shear fractures (Riedel shears). Fault surfaces dip steeply (up to 80°) toward the NNE, whereas Riedel shears are subvertical or dip steeply toward the ENE (see the stereographic projection in Fig. 7D). Slickenlines and abrasive striations on fault planes are characterized by a pitch smaller than 15° or greater than 165° (see the stereographic projection in Fig. 7D). Collectively, the analysis of kinematic indicators (e.g., Riedel shears, lunate-cuspate morphologies on the striated fault planes; Fig. 7E) indicates a right-lateral shear for these E-W–striking faults affecting travertine deposits on both sites. On the other hand, a N-S–striking (N180°/80°) fault crosscuts the fissure ridge at Semproniano village and the eastern edge of the plateau at Poggio Semproniano. At Semproniano village (site 1), this fault develops a 0.5-m-wide damage zone within the banded travertine (Fig. 8A). Slickenlines on the fault surfaces are very difficult to determine. Very close to this N-S–striking fault, centimeters-thick (average thickness: 7–10 cm), NNE-SSW–striking fractures filled by light-brown speleothem-like concretions cut through the banded travertine (Figs. 8B and 8C). Geometrical relationships between the fault surface and these speleothem-filled fractures (here interpreted as tension fractures) suggest an oblique shear for this fault. This N-S–striking fault is probably the northern culmination of the extensional fault system that marks the contact between the Pliocene–Quaternary sedimentary deposits and the Tuscan nappe units south of the village of Semproniano (Fig. 3A).

Faults affecting the host rocks of the travertine deposits are not well exposed in the study area. NW-SE–striking faults have been recognized affecting the Pleistocene continental deposits exposed to the southeast of Semproniano (site 12). These faults consist of fault segments (Fig. DR1B) with spacing of a few decimeters and displacement of a few centimeters at the most. When present, fault striations show a pitch around 160° (Fig. DR1B), indicating oblique strike-slip kinematics. Subsidiary Riedel shears are consistent with right-lateral fault motions. Furthermore, this fault set is aligned along the strike of a main strike-slip fault zone exposed in an abandoned quarry within the Diaspri Formation (site 11; Fig. DR1C). This NW-SE–striking fault zone is characterized by near-vertical fault surfaces, forming a negative flower structure, close to which bedding is severely undulated and verticalized.

At Poggio Semproniano (site 13), calcite-filled veins occur as a systematic set cutting through the bedded travertine and the underlying Scaglia Toscana Formation. These veins are 2–5 cm thick, with variable spacing between ∼0.1 m and 1 m (Fig. 9A). The veins consist of monogenic filling of white crystalline calcite or rhythmic millimeter-thick white-and-gray layering (Figs. 9B and 9C). At Poggio Semproniano, this vein set strikes NW-SE (average strike: N129°) and dips steeply (>70°) toward the SW (Fig. 9A).

At Semproniano (site 1), Poggio i Piani (site 2), Poggio Semproniano (site 4), and Pian di Palma (site 10), travertine deposits are affected by networks of fractures, some of which are strongly karstified. Fractures consist of decimeter-wide mechanical discontinuities within the travertine beds with either curvilinear or planar geometries. Fractures are commonly identified around (several meters from) the main fault systems cutting through the travertine bodies. At Poggio Semproniano and Poggio i Piani, these features strike from ESE-WNW to ENE-WSW, with dip angles close to 90° (Fig. DR1D). Evidence of karst weathering is common along the fracture surfaces, which consist of central cracks surrounded by several secondary anastomosing irregular fractures. The edges of these fractures are often karstified with a typical jigsaw profile. In the I Pianetti quarry, the karst fractures strike NE-SW. These features are meters to decameters long, with a mean spacing of 5–10 m (Fig. 9D). Karstified fractures are often connected, upward or downward, with large subhorizontal cavities. Locally, these fractures host speleothems formed by sparry and globular calcite (Fig. 9E).

At Saturnia (site 8), an exposed steeply dipping, 0.5-m-thick banded travertine cuts through the subhorizontal travertine beds (Fig. 10A). The banded travertine forms an inclined (∼33°) tabular body striking N282° and consists of fine-grained white carbonate concretions forming centimeter-thick bands parallel to the contact with the host travertine beds (Fig. 10B). Similar geometrical and crosscutting relationships have been observed within the travertine exposed to the north of the village of Saturnia (site 7), where a meter-thick banded travertine characterized by undulating bands cuts through the bedded travertine (Fig. DR1E).

Eighteen banded and bedded travertine samples, calcite-filled veins, and speleothem-like concretions were analyzed to constrain the age of hydrothermal circulation in the Albegna basin. The samples were analyzed at the Laboratory of Environmental and Isotopic Geochemistry (Department of Sciences, Roma Tre University, Italy). Samples were cut with a diamond saw and ultrasonically washed in deionized water. About 60 g of each prepared sample were dissolved in 7 N HNO₃ and filtered to separate leachates from insoluble residue. The leachates were heated to 200 °C after adding a few milliliters of hydrogen peroxide to annihilate organic matter, and then spiked with a ²²⁸Th/²³²U tracer. The isotopic complexes of U and Th were extracted according to the procedure described in Edwards et al. (1987) and then analyzed through alpha-counting using high-resolution ion-implanted Ortec silicon-surface barrier detectors. For samples with a ²³⁰Th/²³²Th activity ratio higher than 80 (free from nonradiogenic ²³⁰Th), ages were determined using the measured ²³⁰Th/²³⁴U and ²³⁴U/²³⁸U activity ratios. Sample ages characterized by a ²³⁰Th/²³²Th activity ratio less than or equal to 80, indicating the presence of nonradiogenic (detrital) ²³⁰Th, required a correction based on the assumption of an average ²³⁰Th/²³²Th activity ratio of 0.85 ± 0.36 for all detrital Th (Wedepohl, 1995). All ages were finally calculated using Isoplot (Ludwig, 2003) with errors expressed as ±1σ.

Our ²³⁰Th/²³⁴U geochronological data are reported in Table 3. All samples show low U concentrations ranging from 3 to 208 ppb, with a ²³⁴U/²³⁸U activity ratio between 0.978 and 1.585. The ²³⁰Th/²³²Th activity ratios, which indicate the extent of detrital contamination in the analyzed samples, range between 1.2 and 289. Determined ages span within the 33–214 ka interval. The youngest ages were obtained for bedded travertine at Pian di Palma quarry (33 ± 4 and 49 ± 15 ka) and Bagni di Saturnia (40 ± 7 ka), and for a calcite vein at Poggio Semproniano (39 ± 4 ka). The oldest ages were obtained for the bedded travertine at Semproniano village (214 +50/–37 ka), Poggio I Piani (198 ± 18 ka), and Poggio Semproniano (171 ± 19 ka). Moreover, three ages were older than the limit (ca. 350 ka) of the dating method. These ages are those of the banded travertine at the Semproniano village and I Vignacci localities (Table 3).

Carbon- and oxygen-isotope (δ¹³C and δ¹⁸O) analyses on 38 samples were performed to constrain the chemistry of the parental hydrothermal fluid. Isotopic compositions were measured according to the carbonate-specific method described in detail in Breitenbach and Bernasconi (2011). Approximately 100 µg of powder were placed into 12 mL Exetainers, (Labco, High Wycombe, UK) and flushed with pure helium. The samples were reacted with 3–5 drops of 100% phosphoric acid at 70 °C with a Thermo Fisher GasBench device connected to a Thermo Fisher Delta V mass spectrometer. The average long-term reproducibility of the measurements (based on replicated standards) is ±0.05‰ for δ¹³C and ±0.06‰ for δ¹⁸O. The instrument was calibrated with the international standards NBS19 (δ¹³C = 1.95‰ and δ¹⁸O = −2.2‰) and NBS18 (δ¹³C = −5.01‰ and δ¹⁸O = −23.01‰). Results are expressed in the conventional delta notation (in ‰) against the Vienna Peedee belemnite (VPDB) standard for both δ¹³C and δ¹⁸O. Formation temperatures of the travertines were determined with the equation of Kele et al. (2015), which is based on travertine vent and pool samples. Calculations were run using the present-day δ¹⁸O value of the Bagni di Saturnia hydrothermal spring (−6.4‰ Vienna standard mean ocean water [VSMOW]), which is characterized by a constant temperature of 37 °C (e.g., Minissale, 2004). The oxygen-isotope composition of the travertine precipitating water is therefore assumed to have been similar to that of the presently active spring at Bagni di Saturnia. Stable-isotope composition and calculated paleotemperatures are reported in Table 4 and Figure 11.

Selected samples include banded travertines, bedded travertines, calcite-filled veins, and speleothem-like concretions. With the exception of light-brown speleothem-like concretions in secondary fractures occurring at Semproniano (SPV1 and SPV2; Fig. 8C), all travertines and associated mineralizations showed positive δ¹³C values between 2.8‰ and 10.5‰ (mean value 6.7‰), indicating a thermogene origin (Pentecost, 2005). The δ¹⁸O values are between −12.7‰ and −5.1‰ (VPDB), with a mean value −8.64‰ (Figs. 11A and 11B).

Calculated paleotemperatures range between a minimum of 22 °C and a maximum of 60 °C (Table 4). Samples characterized by the highest calculated paleotemperatures belong to Semproniano village fissure ridge banded travertines, whereas the lowest ones belong to bedded travertines from Pian di Palma quarry. This evidence attests to a trend of decreasing paleotemperature moving from highest to lowest elevations and from north to south in the study area (Figs. 11C and 11D). Banded travertines are generally characterized by mean calculated paleotemperatures of around 10 °C higher than the associated bedded travertine.

The main travertine deposits within the Albegna basin are aligned N-S (Fig. 3A) and are located along major N-S–striking faults or at fault intersections (at Semproniano village, Poggio Semproniano, Poggio I Piani). Our structural analysis documents that the travertine deposits of Poggio Semproniano and Poggio Pancotta are crosscut by right-lateral strike-slip faults oriented E-W and WNW-ESE, respectively. In addition, we documented a N-S–striking fault at Semproniano village that cuts through the fissure ridge with an oblique-extensional shear. This fault pattern of N-S– and E-W–striking faults is similar to nearby hydrothermal areas in southern Tuscany (e.g., Brogi et al., 2010a, 2012; Rimondi et al., 2015) and along the Tyrrhenian margin (see following).

Our geochronological data allow dating of the tectonic-hydrothermal process. The oldest travertines and veins are older than the limit of the dating method (ca. 350 ka), whereas the youngest inactive deposit, at the Pian di Palma quarry, is ca. 33 ka in age. The Albegna basin hydrothermal system and associated travertine deposition are still active at Bagni di Saturnia. Integrating previous data from nearby similar hydrothermal settings (Taddeucci and Voltaggio, 1987; Brogi et al., 2012; Rimondi et al., 2015), we can link the middle Pleistocene onset of hydrothermal activity in southern Tuscany with the main phases of crustal uplift and emplacement of magmatic bodies in the region (Barberi et al., 1994), with the volcanic activity of the Mount Amiata (300–190 ka; Cadoux and Pinti, 2009) and Vulsini Mountains (590–127 ka; Nappi et al., 1995), and with hypogenic speleogenesis (69–19 ka; Piccini et al., 2015). This hydrothermal scenario is consistent with the occurrence of a regional thermal anomaly below the Tuscan magmatic province, producing long-lived convective circulation of endogenic fluids during the Pleistocene (Minissale, 2004).

The relationships between travertine deposition and development of fault-fracture systems can be used to constrain the minimum age of faulting in the region. The E-W–striking right-lateral fault at Poggio Semproniano cut through the bedded travertine dated at 171 ± 19 ka, whereas the N-S–striking fault cut through the banded travertine at Semproniano fissure ridge dated as >350 ka. Tension fractures at the Semproniano fissure ridge are filled by speleothem-like concretions dated 97 ± 9 ka. The karstified fractures at Poggio I Piani and I Pianetti cut through bedded travertine dated 198 ± 18 ka and 107 ± 15 ka, respectively. These ages constrain the tectonic activity in the Albegna basin to the late Pleistocene.

In synthesis, our structural data show that, while travertine deposition within the study region is still active, fossil travertine deposits as young as 170 ka are crosscut by faults. All together, this evidence indicates that travertine deposition can be considered as syntectonic at the regional (basin) scale. If integrated with ages from recent studies (e.g., Brogi, 2008; Brogi et al., 2010a, 2014), our U-Th data show that faulting in southern Tuscany is significantly younger than previously thought and has worked simultaneously with long-lived convective circulation of hydrothermal fluids in the Tuscan magmatic province.

In our data set (Table 4; Fig. 11), only the speleothem-like concretions sampled in some fractures (SPV1 and SPV2) are characterized by negative δ¹³C values of −9.4‰ and −9.7‰, respectively. These negative values are probably due to the high contribution of atmospheric and soil-derived CO₂, suggesting carbonate mineralizations by percolation of meteoric waters within cracks (Pentecost, 2005). This suggests that the Semproniano fissure ridge was probably fully formed at 97 ka, which is the time of deposition of the analyzed speleothem-like concretions. On the other hand, the travertine and veins have positive δ¹³C values (Table 4), indicative of mixing of deep magmatic fluids with meteoric waters having CO₂ originating from limestone decarbonation (Gonfiantini et al., 1968; Guo et al., 1996). Our δ¹³C and δ¹⁸O values are comparable with those reported in previous studies (e.g., Minissale, 2004, and references therein) and are in the range typical of thermogene travertines deposited by present-day thermal springs of central Italy (Minissale, 2004; Gandin and Capezzuoli, 2008). Travertines belonging to Semproniano village fissure ridge show δ¹³C values more positive than usual thermogene values, likely attributable to downstream CO₂ degassing, producing an increase in δ¹³C (Özkul et al., 2013).

Travertine deposits of the Albegna basin are hydrothermally distinct from nearby hydrothermal deposits along the Tyrrhenian margin. Continental carbonates of the Sarteano system (southern Tuscany) yielded δ¹³C values ranging between −2.5‰ and 1.6‰ (Brogi et al., 2012), indicative of a larger meteoric component. The Tivoli travertine (Latium) shows a range of δ¹³C between 8.31‰ and 10.77‰ and a range of δ¹⁸O between −4.76‰ and −7.18‰, which were attributed to a strong process of diagenesis that obliterated the original oxygen-isotopic signature (Manfra et al., 1974; De Filippis et al., 2013a). The variability of oxygen- and carbon-isotope compositions observed in the travertines of the Albegna basin has analogies with those of travertine samples from the Denizli basin (Turkey) and Mammoth Hot Spring (Yellowstone Park, USA). For the Denizli basin, the travertine formation is largely attributed to variable interaction between the meteoric waters and the deep hydrothermal fluids (e.g., Dilsiz, 2006; Uysal et al., 2007; De Filippis et al., 2013a). In contrast, the travertine deposition at Mammoth Hot Spring is attributed to a high extent of CO₂ degassing during diagenetic processes (e.g., Fouke et al., 2000; Chafetz and Guidry, 2003). Based on carbon- and oxygen-isotope data, travertines of the Albegna basin can be interpreted as forming during dominant circulation of CO₂-enriched hydrothermal fluids with variable contributions of colder shallow aquifer fluids mainly recharged by meteoric precipitation.

Estimated fluid temperatures span from ∼60 °C for the banded travertine from Semproniano to ∼22 °C for the bedded travertine from the Pian di Palma quarry (Table 4; Figs. 11C and 11D). There are no analogous data for fossil hydrothermal deposits along the Tyrrhenian margin in central Italy, but our estimated temperatures are comparable with the temperatures of active thermal springs in this region (temperature > 20 °C; Minissale, 2004).

The longevity of individual travertine deposits and structures in the Albegna basin cannot be constrained with the available age data, but the activity of the entire Albegna hydrothermal-travertine system must be longer than 350 k.y. This is more than estimated activities at Rapolano (Tuscany, Italy, ∼133 k.y.; Brogi et al., 2010a), at Sarteano (Tuscany, Italy, ∼250 k.y.; Brogi et al., 2012), at Tivoli (Latium, central Italy, ∼100 k.y.; Faccenna et al., 2008), at Limagne graben, France (Massif Central, ∼250 k.y.; Rihs et al., 2000), and in the Ebro basin, Spain (∼239 k.y.; Luque and Julià, 2007). Longer activities on the other hand, have been documented for travertine deposition in the Denizli basin, Turkey (at least 600 k.y.; Engin et al., 1999; Altunel and Karabacak, 2005; Uysal et al., 2007; De Filippis et al., 2013a,Özkul et al., 2013; Lebatard et al., 2014; Toker et al., 2015), and in the Rio Grande rift, United States (∼660 k.y.; Priewisch et al., 2014).

We can interpret the travertine deposits of the Albegna basin as markers of hydrothermal activity that evolved in space and time along a N-S structural alignment. The following main features have been recognized and documented:

• (1) There is a general southward rejuvenation of the hydrothermal system, i.e., the travertine deposition becomes younger moving from north to south. The Semproniano deposits have ages older than 350 ka, whereas the youngest dated deposit (ca. 40 ka) and the active springs occur toward the south in the Bagni di Saturnia locality.

• (2) There is a decrease in travertine deposition from 600–700 m a.s.l. at Semproniano and Poggio Semproniano to 140–220 m a.s.l. at Bagni di Saturnia and I Pianetti. The parallel elevation and chronological gradients (from north to south) indicate a rapid lowering of the water table toward the south (i.e., toward the Tyrrhenian Sea) during Pleistocene time on the order of 1 mm yr^–1 (Piccini et al., 2015).

• (3) There is a change in isotopic signatures. Along the N-S profile (Fig. 11A), a general decreasing trend of δ¹³C and increasing trend of δ¹⁸O from older to younger travertine deposits (Fig. 11B) are observed. The only exceptions are the calcite-filled veins occurring at Poggio Semproniano, which developed synchronously with the bedded travertine of Bagni di Saturnia and Pian di Palma localities. The δ¹⁸O values show a wide range, with the most negative values corresponding to older travertines (Semproniano village, Poggio Semproniano, and Poggio I Piani) and the less negative values belonging to the travertines of Pian di Palma quarry (the youngest deposit). The gradient of δ¹⁸O values from north to south is interpreted as a decrease in temperature (Fig. 11C). This decreasing trend is also confirmed when considering the temperatures versus elevations of the travertine deposits (Fig. 11D). This evidence supports the hypothesis of an increasing contribution of meteoric waters with time.

Summarizing, we interpret the Albegna basin as a long-lived hydrothermal-tectonic setting with a progressive southward migration of the fluid circulation, moving away from the geothermal center of Mount Amiata, located to the north of the Albegna basin (Fig. 12). The main travertine-depositing center changed its position horizontally (along the N-S structural alignment) and vertically (lowering of the depositional elevations), reaching the present-day deposition center at the Bagni di Saturnia locality. The travertines show decreasing δ¹³C and increasing δ¹⁸O consistent with increasing dilution of endogenic fluids by meteoric fluids. The main extensional and strike-slip fault systems provided the necessary hydraulic pathways for infiltration of meteoric waters to the depth of the carbonate reservoir, the possible connection and mixing between different hydrogeological-hydrothermal circuits (e.g., Curewitz and Karson, 1997; Cox et al., 2001; Rowland and Sibson, 2004; Gudmundsson, 2011), and the subsequent ascension to surface of CaCO₃-rich hydrothermal fluids. Mixing with meteoric fluids resulted in a decrease of fluid temperatures down to the present value of 37 °C at the Bagni di Saturnia springs (Minissale, 2004).

Our geochronological data can contribute to the understanding of the feedback relationships between thermogene travertine deposition and paleoclimate in the Albegna basin. In the last decades, many studies have been devoted to the correlation between travertine deposition and paleoclimate oscillations. There is a general consensus in considering warm and wet (interglacial) conditions as the most favorable for travertine deposition during late Quaternary time (e.g., Dramis et al., 1999; Frank et al., 2000; Rihs et al., 2000; Soligo et al., 2002; Pentecost, 2005; Luque and Julià, 2007; Faccenna et al., 2008; Kampman et al., 2012; Priewisch et al., 2014). Intuitively, highstand conditions of the water table during wet periods can favor the supply of fluids for the growth of travertine deposits (Rihs et al., 2000; Faccenna et al., 2008). Nevertheless, travertine formation in dry glacial periods (lowstand of the water table) has been documented and used to emphasize the importance of tectonic activity, rather than climate, to control travertine precipitation (e.g., Uysal et al., 2009; Brogi et al., 2010a; Özkul et al., 2013). Finally, complete interaction and feedbacks among fluid discharge, paleoclimate, and tectonics has been proposed by De Filippis et al. (2013a) to morphologically and volumetrically control the different travertine deposits (travertine plateaus and fissure ridges).

We compared the age of travertine deposition (Table 3) with major Quaternary paleoclimate indicators and events determined both at the global and regional scales (Fig. 13). We used, in particular, paleoclimate records extracted from the deep-sea oxygen-isotope trend (Zachos et al., 2001) integrated with the pollen data set from Valle di Castiglione (located ∼200 km to the south of the Albegna basin; see location in Fig. 1). Although the travertine ages have rather large error bars, the majority of them fall within interglacial or interstadial periods. This suggests that the travertines and travertine-related structures preferentially formed during warm (and humid) climate periods characterized by highstand conditions of the water table. Nevertheless, our data also document that large volumetric amounts of travertine deposits in the Albegna basin (bedded travertine plateaus at Poggio Semproniano and at Saturnia village) were formed coeval with a glacial period at ca. 130–180 ka, which is also a nonhumid time, as demonstrated by the pollen curve from Valle di Castiglione (Fig. 13). This evidence probably indicates an important contribution of the endogenic fluid supply to travertine deposition during the lowstand (and nonhumid) conditions of the water table (see also Toker et al., 2015). In other words, the endogenic fluid supply may have partly compensated for the lowering of the water table during the glacial lowstand periods.

The recognized structural pattern of the Albegna basin is compatible with the postorogenic regional extensional tectonic regime described for the Tyrrhenian margin, where main NW-SE–striking extensional faults (i.e., parallel to the trend of the Apennines belt) and related sedimentary basins developed in Pliocene–Quaternary times (Malinverno and Ryan, 1986; Patacca et al., 1990; Bossio et al., 1993; Martini and Sagri, 1993; Bartole, 1995; Jolivet et al., 1998; Cavinato and De Celles, 1999; Pauselli et al., 2006; Billi and Tiberti, 2009; Brogi et al., 2014). Within this scenario, the strike-slip to transtensional faults oriented transversally to the NW-SE extensional boundary faults are interpreted as transfer systems accommodating different stretching rates within the extending crust (Faccenna et al., 1994; Aiello et al., 2000; Acocella and Funiciello, 2002, 2006; Liotta et al., 2015). It is worth nothing that many of these transverse faults have acted as preferential pathways for hydrothermal outflow and controlled distribution of volcanism along the Tyrrhenian margin at least since late Pleistocene time. In this regard, we report, from NW to SE along the Tyrrhenian margin, 10 main instances (see Fig. 14A) where recent or active hydrothermal activity is localized along transverse faults (i.e., NE-SW and N-S), as is the case of the Albegna basin (this study): (1) Rapolano, Tuscany (Brogi and Capezzuoli, 2009), (2) Sarteano, Tuscany (Brogi et al., 2012), (3) Torre Alfina, Latium (Vignaroli et al., 2013), (4) Roman Plain, Latium (Sella et al., 2014; Bigi et al., 2014; Frepoli et al., 2010), (5) Viterbo, Latium (Baldi et al., 1974; Chiocchini et al., 2010), (6) Sutri, Latium (Corrado et al., 2014), (7) Tivoli, Latium (Gasparini et al., 2002; Faccenna et al., 2008; De Filippis et al., 2013a, 2013b), (8) Zolforata di Pomezia, Latium (Vignaroli et al., 2015), (9) Zannone Island, Latium (Ingrassia et al., 2015), and (10) Mount Massico–Roccamonfina, Campania (Billi et al., 1997; Corniello et al., 2015).

Transfer faults are known as compartmental non-Andersonian faults restricted to shallow levels, i.e., in the hanging wall of extensional faults (Gibbs, 1984; van der Pluijm and Marshak, 2003). As such, transfer faults are not expected to act as pathways for endogenic (deep) hydrothermal fluids as is the case of the Albegna basin and elsewhere along the Tyrrhenian margin. We conclude that the structures that nucleated and initially grew as transfer zones oriented transversally to the Apennines trend (Neogene) have probably evolved as through-going crustal faults in recent times (Quaternary) so as to allow the upflow of endogenic fluids stored in geothermal reservoirs of the region. (1) In a first stage (Neogene; Fig. 14B), the postorogenic extensional tectonics that acted on the Tyrrhenian margin of the Apennines operated through major NW-SE–striking extensional faults. The orientation of the paleostress axes followed Anderson’s theory of extensional faulting with a vertical σ₁ (maximum compression) direction and σ₂ and σ₃ (intermediate and minimum compression, respectively) directions lying on a horizontal plane. The vertical σ₁ direction was the consequence of far-field (slab retreat) and near-field (Apennines crustal thickening) stress regimes acting at the back of the eastward-migrating compressional system. Within this setting, transfer zones developed in a non-Andersonian mode in the hanging walls of the NW-SE–striking normal faults to separate differently stretched adjacent compartments (e.g., Gibbs, 1984). (2) As the extension progressed (late Quaternary; Fig. 14B), vigorous crustal thinning and erosion reduced the lithostatic load (σ₁), providing a new stress regime probably characterized by a vertical σ₂ axis (switch between σ₁ and σ₂) and horizontal σ₁ and σ₃. At this stage, the NW-SE–striking extensional faults became less effective, whereas the former shallow transfer zones switched to strike-slip crustal-scale faults. Both the preexisting NE-SW–striking and some newly generated N-S–striking strike-slip to transtensional faults responded to a horizontal σ₁ and propagated downward into the crust so as to reach the endogenic fluids accumulated at depth and offer them a viable pathway for surficial uprising and mixing with meteoric waters.

• (1) Quaternary travertine deposition in the Albegna basin in central Italy reflects the interaction and feedbacks among hydrothermal activity, active tectonics, and paleoclimate within a region of positive geothermal anomaly. While hydrothermalism provided the fluids and the heat, tectonics controlled the location of faults, and paleoclimate modulated the abundance and elevation of groundwaters that buffered and mixed with the endogenic fluids.

• (2) The reason for the southward and downward migration of the thermogene travertine depositional system remains to be explained. As such, the Albegna basin remains an interesting topic to be further studied for a better understanding of additional causes, e.g., eustatism, morphological molding, and groundwater-level changes as influenced by climate oscillations and endogenic fluid supply.

• (3) The reconstructed spatio-temporal tectonic-hydrothermal evolution of the Albegna basin sheds light on the neotectonic and hydrothermal activity along many faults transverse to the Tyrrhenian margin, where, due to a change in the stress tensor induced by postorogenic crustal thinning, these originally shallow faults were reactivated and enhanced so as to propagate rupture downward into the crust and generate effective pathways for hydrothermal outflow.

We thank C. Faccenna for fruitful discussions and G. Fellin for her help in the field and in the laboratory at ETH Zürich. The “Saturnia Travertini” company is acknowledged for kindly providing quarry access and logistic assistance. S. Kele was supported by the János Bolyai scholarship of the Hungarian Academy of Sciences and the Hungarian Scientific Research Fund (OTKA 101664). This manuscript benefited from comments and advice from S. Zanchetta and two anonymous reviewers. The editor (K. Stüwe) is acknowledged for the editorial handling.