Abstract
This paper is based on a multidisciplinary field and laboratory study of carbonate concretions developed in poorly lithified Quaternary, syn-kinematic sediments along the Quattro Castella Anticline, Northern Apennines, Italy. The studied concretions consist of both tabular (parallel to bedding) and elongate single to coalescent concretionary bodies oriented at different angles to the bedding throughout the exposed stratigraphic succession. The dimensions of the concretions range from a few centimetres for single elongate concretions up to several metres for tabular to coalescent concretions. Field observations and petrophysical data indicate that the concretions developed preferentially in sediments characterized by mean grain sizes of 90–290 μm and a permeability ranging from 7 × 102 to 7 × 104 mD. Carbon and oxygen stable isotope analyses in conjunction with the petrographic investigations indicate that the precipitation of concretionary calcite occurred in a meteoric vadose realm during early eogenesis and subsequently in a meteoric phreatic environment. Diagenetic data and concretion patterns in syn-tectonic sediments suggest they formed during the lateral propagation of the anticline, which, in turn, promoted a change in the local topographic–hydraulic gradient from fold axial parallel to fold orthogonal. The integrated analysis of carbonate concretions provides a useful tool with which to unravel the palaeo-fluid flow history and therefore to predict fluid circulation patterns in folded siliciclastic rocks.
Supplementary material: Complete isotopic data from both hand samples and thin sections, together with statistical analysis, scanning electron microscope cement images and grain size distributions with the standard operative procedure tests performed, and detailed permeability measurements are available at https://doi.org/10.6084/m9.figshare.c.4087019
Unravelling fluid flow pathways within exposed rocks and sediments leads to a better understanding of the processes affecting oil and gas reservoir quality and facilitates hydrocarbon exploitation and the preservation of groundwater reserves (Taylor et al. 2000; Dutton et al. 2002; Hall et al. 2004). Diagenetic concretions and mineral masses can provide a useful tool with which to reconstruct palaeo-fluid flow patterns within transforming porous media involved in deformation (e.g. Johnson 1989; McBride et al. 1994; Mozley & Davis 1996; McBride & Parea 2001; Balsamo et al. 2012, 2013). Moreover, selective cementation plays a key part in modifying the petrophysical properties of sediments, such as porosity and permeability (Houseknecht 1987; Davis et al. 2006; Boutt et al. 2014). Pervasive cementation can eventually lead to the formation of laterally extensive, bedding-parallel, low-permeability layers, which can compartmentalize reservoirs (Morad et al. 2000; Dutton et al. 2002; Hall et al. 2004). Diagenetic concretions within siliciclastic sediments are characterized by different shapes and sizes (e.g. Sellés-Martínez 1996; Seilacher 2001) depending on the fluid flow patterns and the spatial distribution of sediments (McBride et al. 1994; Davis 1999; McBride & Parea 2001; Cavazza et al. 2009; Balsamo et al. 2012).
The arrangement of concretions has been used as marker of fluid flow and to recognize the diagenetic setting in which precipitation occurred (Pirrie 1987; McBride et al. 1994; Mozley & Davis 1996; Davis 1999; South & Talbot 2000). Different modes of growth have been identified based on the type of concretion (Mozley 1996; Abdel-Wahab & McBride 2001; Mozley & Davis 2005). A significant amount of research has been carried out on the relationship between diagenetic concretions and fault zones within siliciclastic sediments with different degrees of lithification (Mozley & Goodwin 1995; Hendry & Poulsom 2006; Balsamo et al. 2012, 2013; Philit et al. 2015; Williams et al. 2016). However, comparable information is lacking on the presence and patterns of carbonate concretions hosted in syn-kinematic sediments deformed in growing folds, where the pattern of fluid flow may change during fold amplification.
We present here the results of a multidisciplinary study on diagenetic concretions developed within poorly lithified Quaternary syn-kinematic sediments exposed in the Quattro Castella Anticline at the mountain front of the Northern Apennines. The petrophysical and sedimentological properties of the sediments hosting the concretions were studied by laboratory analyses (grain size distribution, two-dimensional porosity) and in situ measurements (permeability). We inferred the diagenetic environment and the nature of the fluids responsible for the formation of concretions using stable isotope analysis combined with petrographic observations of carbonate cements. The integration of field and laboratory data allowed us to constrain the control of the lateral propagation and growth of the anticline on the palaeo-fluid flow pathways.
Geological framework
The Apennines fold–thrust belt developed from early Oligocene times as a consequence of the subduction of the Adria plate below the European plate (Boccaletti et al. 1971; Doglioni 1991). Subduction was responsible for the emplacement of remnants of oceanic crust and the related sedimentary cover of the Ligurian–Piedmont ocean (Ligurian Units) onto the inverted Adria passive margin successions (Doglioni 1991). The Epiligurian wedge-top basin successions (Eocene to Late Miocene) formed on top of the Ligurian thrust sheets during their northeastwards migration (Zattin et al. 2002; Remitti et al. 2011; Piazza et al. 2016).
The study site is located at the mountain front of the Northern Apennines (Fig. 1a) and is characterized by Ligurian and Epiligurian rocks, which were overthrust onto Miocene siliciclastic foredeep turbiditic sediments during Middle Pliocene times (Oppo et al. 2013, 2015; Gunderson et al. 2014). The Quattro Castella Anticline formed in the hanging wall of a blind thrust ramp cutting through the foredeep strata, which were previously overthrust by a Ligurian thrust sheet and the associated Epiligurian sediments (Fig. 1b). Such a blind thrust has been active since Late Miocene times and is still active today (Ponza et al. 2010; Boccaletti et al. 2011; Gunderson et al. 2014). The stratigraphic succession investigated in this work lies in the forelimb of the WNW–ESE-striking Quattro Castella Anticline, a regional-scale fold that deforms the Ligurian and Epiligurian rocks and part of the Po plain autochthonous succession (Fig. 1a and b). In particular, remarkable and continuous exposure is provided by the erosional incision of the Enza River (Fig. 2a). The outcrop consists of Pleistocene and Holocene sediments belonging to the Po plain succession that were deposited in the western periclinal termination during the late stage of fold growth. The outcrop may have recorded a different pattern of fluid flow induced by the anticline lateral amplification. From the base, it is composed of the silty clay Argille Azzurre Formation deposited in a shallow prodelta environment. It is followed by the fluvio-deltaic sand of the Sabbie di Imola Formation, overlain by the Lower Emilia–Romagna synthem, which is mainly characterized by fluvio-continental deposits (Gunderson et al. 2014).
(a) Simplified geological map of the topographic front of the Northern Apennines showing the position of the study area along the Enza River valley (modified after Oppo et al. 2015). (b) Reflection seismic section with line drawing interpretation of the main structural features in the vicinity of the studied field site (modified after Oppo et al. 2015).
(a) Detailed geological map of the Enza River outcrop. Stereonets (Schmidt equal area projection lower hemisphere) report the mean bedding planes with dip angles and poles to bedding measured at different sites along the outcrop. (b) Detailed geological section across the entire Enza River outcrop. Bedding dip domains are also reported and show a progressive decrease in dip from the base to the top of the succession.
Analytical methods
A total of 620 structural data values were collected to construct the geological cross-section through the outcrop (Fig. 2). A c. 300 m thick stratigraphic succession was logged, excluding the uppermost 70–80 m thick fluvial conglomerates. A 30 m thick stratigraphic log was measured in greater detail to investigate the typology and arrangement of carbonate concretions. Within this interval, the petrophysical properties of the sediments hosting concretions were measured. In particular, 40 grain size samples were analysed using a Malvern Instruments Mastersizer 3000 laser diffraction particle size analyser with an operating range from 0.1 to 3500 μm, equipped with a Hydro EV wet dispersion unit using de-ionized water as the dispersant fluid. Detailed information about the adopted standard operating procedures are provided in the Supplementary Material. The sediment permeability corresponding to the 40 grain size sampling sites was measured using a New England Research Tiny Perm II air permeameter. The instrument has a rubber nozzle with a central hole of 0.5 cm diameter and the data were calibrated according to the method used by Balsamo et al. (2012). We performed 900 in situ permeability measurements throughout the 30 m thick detailed log. A set of 780 permeability measurements was also measured on the outer surface of the concretionary bodies (see Supplementary Material).
A total of 12 concretions of different sizes and shapes were sampled in different stratigraphic positions along the section. Eleven of the concretions were from the sandy strata of the Sabbie di Imola Formation, whereas only one was sampled within the Argille Azzurre Formation, in the same interval studied by Oppo et al. (2015), and was used for comparative purposes. Hand specimens of the concretions were cut to obtain 14 thin sections impregnated with a blue-dyed resin to calculate the two-dimensional porosity and intergranular volume (IGV) from scanned high-resolution images using ImageJ image analysis software. Polished rock slabs and thin sections were stained with Alizarin Red S and potassium ferricyanide to identify the cement mineralogy (dolomite v. calcite) (Dickson 1966). Cold cathodoluminescence (CITL CL Mk5-2 with operating settings of 250 μA and 10–12 kV) was used to better describe the texture and spatial distribution of the cement. A JEOL JSM 6400 scanning electron microscope (operating settings 240 μA and 20 kV) was used to acquire detailed photomicrographs of the cement.
The bulk rock carbon and oxygen stable isotopes were analysed from powder samples extracted from fresh section cuts with a dental drill along two orthogonal transects with different sample spacings (1 cm for the horizontal trace and 0.5 cm for the vertical trace). A total of 184 samples was collected. The concretion cements were also sampled in 45 different spots identified from thin sections using a New Wave Research micromill with a 300 μm diameter drill bit. Micromill sampling was performed at the Department of Earth and Environmental Sciences, KU Leuven, Belgium. All samples were analysed with a Thermo Finnigan DELTA plus XP mass spectrometer coupled with a Thermo Finnigan GasBench II gas preparation and introduction system. All isotopic analyses were carried out at the Department of Chemistry, Life Sciences and Environmental Sustainability of the University of Parma, Italy. Both δ13C and δ18O are referred to the international standard V-PDB (Pee Dee Belemnite). The analytical precision for the carbon isotope determination was 0.10‰ V-PDB and that for the oxygen isotopes was 0.15‰, whereas the prediction uncertainty was c. 0.15‰ for carbon and c. 0.20‰ for the oxygen isotopes.
Stratigraphic and structural description
Outcrop features
The stratigraphic interval recorded in the 300 m long exposure spans a time period from 1.65 Ma at the base to Holocene deposits at the top (Gunderson et al. 2014). From the base to the top, two different synthems can be identified: (1) the Marine Quaternary synthem; and (2) the Lower Emilia–Romagna synthem (see Gunderson et al. 2014). From a lithostratigraphic point of view, the Marine Quaternary synthem is composed of two formations: the Argille Azzurre Formation and the Sabbie di Imola Formation.
The Argille Azzurre Formation consists of a grey–cyan silt to silty clay, organized in metre-thick strata with weakly defined stratification. The age covered by the formation ranges from 1.65 to 1.1 Ma (Gunderson et al. 2014) with a total thickness of 121 m. It can be further subdivided in two distinct units. The lower unit is a 70 m thick unit composed of relatively uniform silty clay with thin, very fine-grained sand layers containing sparse marine fossils and fossilized methane expulsion vents. The upper unit of the Argille Azzurre Formation has a total thickness of 51 m and includes three distinct fine sand units, each composed of 8–10 m thick fine-grained laminated sand, in which asymmetrical to symmetrical ripples are present, separated by thick silty clay bodies (Fig. 2b). The uppermost part of the second unit shows a progressive increase in the sand fraction. Consequently, the Argille Azzurre Formation is characterized by a general coarsening upwards and the upper coarse-grained interval marks the transition to the overlying Sabbie di Imola Formation.
The Sabbie di Imola Formation is located almost in the middle part of the investigated succession (Fig. 2b). It consists of yellow–orange, medium- to fine-laminated sand with a total thickness of 28 m. The age spans from 1.1 to 0.8 Ma (Gunderson et al. 2014). This formation is composed of two distinct sand units separated by a 2 m thick, grey–blue fossil-rich clay. The basal sand body is made of medium- to fine-grained sand organized in amalgamated strata with hummocky cross-stratification. The total thickness of this unit is 13 m. The second depositional unit of the Sabbie di Imola Formation shows an increase in grain size and is mainly composed of an alternation of medium-grained sand and pebble–cobble orthoconglomerates for a total thickness of 12 m (Fig. 2b). The Lower Emilia–Romagna synthem unconformably overlies the Sabbie di Imola Formation with an age between 0.8 and 0.4 Ma (Gunderson et al. 2014). At the base, it shows an 11 m thick green to grey clay with faint stratification characterized by abundant plant roots and extensive orange–red colour mottling. These deposits are overlain, with a sharp erosive surface, by a pebble to cobble conglomerate, followed by a 4 m thick conglomeratic mudstone. The uppermost portion of this synthem is formed by a sandy unit with a thickness of c. 15 m. The Lower Emilia–Romagna synthem passes upwards, through a well-defined erosional unconformity, into the 70–80 m thick conglomerates body of the Upper Emilia–Romagna synthem (Fig. 2b). Consequently, the stratigraphic succession is characterized by a coarsening and shoaling upwards facies sequence that records a transition from an offshore marine environment (Argille Azzurre Formation) to an alluvial environment (the Lower and Upper Emilia–Romagna synthem) (Gunderson et al. 2014).
Sedimentary wedging
The entire outcrop displays a wedge-shaped geometry with strata pinching towards the south and opening basinwards to the north. The investigated succession was deposited along the forelimb of the growing anticline, thus showing a progressive decrease in the bedding dip angle from the base to the top of the outcrop. In the first few metres of the succession, within the Argille Azzurre Formation, the bedding dip has a maximum value of 55°; going upsection through the basal unit of this formation (i.e. within the first 70 m of the measured stratigraphic log), it decreases to 36° (Fig. 2a and b). This decrease occurs progressively and no sharp unconformity is visible in the field. Within the second unit of the Argille Azzurre Formation, the bedding dip continue to decrease from 36 to 25° at the top of the transition between the Argille Azzurre and Sabbie di Imola formations. In this interval, the variation in the bedding dip is no longer gradual, but occurs along the erosive surfaces at the base of the three fining-up sedimentary cycles. The base of the Sabbie di Imola Formation has an average dip of 25°, which decreases to 23° within the 2 m thick clay layer separating the two sand units (Fig. 2b). Throughout the second sandy unit, the bedding dip remains almost constant at 22–23°. The Lower Emilia–Romagna synthem shows a bedding dip decreasing from 18° in the continental clay to almost horizontal at the top of the sedimentary succession (Fig. 2b). The dip direction of the strata remains almost constant throughout the entire section, with values spanning from N 2° to N 25°.
Carbonate concretions
Concretion types and patterns
The stratigraphic interval in which the diagenetic concretions are densely concentrated includes the transition between the Argille Azzurre and the Sabbie di Imola formations and the first sand body of the latter formation (Fig. 2b). Based on their shape, size and spatial distribution, the concretions can be classified into four categories as follows.
Single elongate, cigar- to blade-shaped concretions occur within the medium–fine sand strata of the first body of the Sabbie di Imola Formation (Fig. 3a). In some sand layers, these concretions are clustered and closely spaced, without merging together.
Elongate coalescent concretions, which make up metre-scale wide bodies, can be found within the basal sand strata of the transition from the Argille Azzurre to the Sabbie di Imola formations and at the top of the first sand unit of the latter formation. These concretions are massive and emerge by their strong positive relief with respect to the host sand due to differential weathering. Concretionary bodies are formed by the joining of several single elongate bodies (Fig. 3b).
Tabular concretions are often distributed within the first sand body of the Sabbie di Imola Formation. They appear as laterally extensive layers, oriented parallel to the sand strata bedding, showing a variable thickness from 2 to 20 cm (Fig. 3c). The degree of cementation is high, even if a few tabular bodies are friable in hand specimen. The majority of these concretionary bodies display a clear elongation direction and they seem to be formed by several single, small-sized elongate concretions merged together. The merging aspect is clearly visible on the basal and upper surfaces of the tabular concretions, but it is not distinguishable in cross-section.
Nodular concretions display almost spherical shapes with diameters from 4 to 15 cm. They are densely clustered in three different layers near the middle of the first sand unit of the Sabbie di Imola Formation. The degree of cementation varies inside each nodule, with some small sand volumes strongly cemented and others almost non-cemented (Fig. 3d).
Types of concretion hosted within the stratigraphic succession. (a) Elongate single cigar- or blade-shaped concretions emerging from loose sand at the base of the Sabbie di Imola Formation. (b) Coalescent elongate concretion displaying elongation traces on the top surface. (c) Tabular metre-wide bedding-parallel concretions. (d) Nodular and tabular bedding-parallel concretions.
The geometrical relationship between the orientation of the concretions, marked by their longest axis, and the bedding of the host sediment varies along the investigated stratigraphic succession (Fig. 4b and c). In particular, at the transition between the Argille Azzurre and Sabbie di Imola formations, the measured elongation direction, both on single and coalescent concretions, shows a mean trend pointing towards the west. The concretion plunge is not parallel to the bedding dip direction (which is towards the NNE) and is roughly oriented along the strike direction of the strata. The plunge values vary from 1 to 19°. Moving upsection to the base of the first sand unit of the Sabbie di Imola Formation, the mean elongation direction displays a NNW orientation (Fig. 4c). Even in this interval, the concretion plunge is not parallel to the bedding dip direction, but describes a narrower angle to it compared with the underlying concretions. The plunge values are similar to the strata dip angle (23°). Near the middle of the first sand unit, elongate concretions show a mean direction pointing almost to the north (Fig. 4c). The concretion plunge is constant and almost parallel to the strata dip direction. At the top of the basal sand unit, elongate concretions plunge towards the NNE, i.e. parallel to the bedding dip direction (21°). From the base to the top of the detailed log (Fig. 4b) (c. 25 m in stratigraphic height), the direction of the elongate concretions describes a 110° clockwise rotation (Fig. 4c).
Stratigraphic, structural and petrophysical properties of the investigated succession. (a) Stratigraphic log of the entire exposed succession. (b) 30 m thick detailed log showing the transition between the Argille Azzurre and Sabbie di Imola formations and the first sand unit of the latter formation. (c) Concretion structural data through the detailed log (Schmidt equal area projection lower hemisphere). (d) Grain size and permeability variation curves.
To recognize the changes in bedding and concretion orientation during sedimentation, unfolding of tilted strata was performed considering the dip domains illustrated in the geological cross-section (Fig. 2b). During the deposition of the strata marking the transition between the Argille Azzurre and Sabbie di Imola formations, the bedding was almost sub-horizontal, with a slight dip towards the west (Fig. 5). With ongoing sedimentation, the bedding dip direction recorded a progressive clockwise rotation from the west to the NE. The same rotation is also displayed by the sand strata of the Sabbie di Imola Formation (Fig. 5). Unfolding of the mean direction of elongate concretions indicates that the diagenetic bodies were only affected by a coaxial rotation during the tilting of the strata. This is confirmed by the increase in dip values and by the almost constant strike direction characterizing the longest axis of the concretions during the deposition and tilting of the host strata.
Progressive tilting of the strata calculated from the unfolding of the dip domains shown in Figure 2b within the stratigraphic interval consisting of the Argille Azzurre–Sabbie di Imola (transition AZ–SI) and the Sabbie di Imola Formation. The mean direction of the elongate concretions was calculated in the same deformation stages and is displayed as the azimuth and dip angles (Schmidt equal area projection lower hemisphere).
Petrophysical properties
Within the transition from the Argille Azzurre to the Sabbie di Imola formations, the sand strata hosting the concretions have a mean grain size varying between 145 and 290 μm with moderate sorting. The mean permeability of these strata spans from 1.6 × 104 ± 3.4 × 103 to 4.2 × 104 ± 5.6 × 103 mD. By contrast, the grain size and permeability show a reduced variability range within the Sabbie di Imola Formation. The sediments hosting the concretions have a narrow grain size range varying between 103 and 188 μm with high sorting, whereas the permeability ranges from 1.2 × 103 ± 2.3 × 102 to 2.3 × 104 ± 3.1 × 103 mD (Fig. 4d). Concretions did not develop below a grain size of 90 μm and a permeability of 600–700 mD.
The porosity is almost completely interparticle, showing a patchy distribution with rare intraparticle and partial mouldic porosity characterizing only fossil shells affected by partial dissolution. The two-dimensional concretion porosity ranges between 1.7 and 4.2% (Table 1). Only one elongate concretionary body displays an anomalously higher porosity of 16.9%, which lies outside the range identified by other concretions (Table 1). The IGV in the sampled concretions varies from 44.9 to 54.8% (Table 1).
Morphology, main cement types, percentage porosity, cement and intergranular volume of the sampled concretions
Concretion petrography
The sampled concretions are composed of abundant quartz, K-feldspar and plagioclase grains, with minor amounts of opaque iron oxides, a few biotite and muscovite micas, rare lithic fragments and pyrite (Fig. 6a, c, e and g). Apart from the calcite cement, the carbonate phases are mostly represented by several species of foram, mollusc and gastropod shell fragments, as well as crinoids, peloids and a few detrital calcite grains.
Photomicrographs of the cement on stained thin sections under transmitted light and under cathodoluminescence. (a, b) Circumgranular rim of equant calcite (Cr) around bioclast followed by blocky zoned calcite (C1) with several zonations inside a tabular concretion sample. (c, d) Rare meniscus dully luminescent calcite (Cm) and fine equant sparitic calcite (C2) surrounding bioclast inside single elongate concretion. (e, f) Syntaxial overgrowth around crinoid fragments and blocky zoned calcite (C1) in tabular concretion. Circumgranular-isopachous rim of equant calcite (Cr) is also present around bioclast. (g, h) Fine equant sparite (C2) post-dating blocky zoned calcite (C1) inside elongate coalescent concretion. B, bioclast; K-f, feldspar; P, porosity; Qz, quartz.
The carbonate cement inside the concretions is made of non-ferroan calcite (pink staining colour) (Fig. 6a, c, e and g). The majority of the cement is pore-filling, with a lesser amount being a grain-lining and coating cement; the cement therefore reduces the interparticle porosity.
Based on the petrographic analysis, four main cement types can be distinguished.
Cr: circumgranular–isopachous rims of small-sized equant calcite (crystal size <10–20 μm) around bioclasts and fossil shells. This cement has a dark cathodoluminescence response (Fig. 6b and f) and does not contain fluid inclusions.
Cm: pore-lining and pore-filling fine sparitic calcite (crystal size <25 μm) with a mosaic texture, located around quartz and feldspar grains and forming a meniscus cement. It is dominantly non-luminescent, but in a few areas a bright orange luminescence occurs (Fig. 6d). No fluid inclusion is present inside this cement.
C1: coarse–blocky sparite with a drusy texture (the size of the cement crystals increases towards the centre of the pore), together with syntaxial overgrowths around crinoid fragments, biogenic and detrital calcite (Fig. 6f). This cement is characterized by crystal sizes up to 200–300 μm and the majority show a zoned cathodoluminescence pattern consisting of alternating non-luminescent and bright sub-zones (Fig. 6b and h). The sub-zone number and width vary according to the type of concretion. Tabular concretions display the highest number of sub-zones with smaller widths, whereas elongate single concretions have just a few, thick, less-defined sub-zones. Other crystals belonging to the same cement pattern do not display any zonation and are non-luminescent. Syntaxial overgrowths may develop crystals up to 400 μm. They are characterized by less bright sub-zones than the blocky calcite and the crystal cores are usually non-luminescent (Fig. 6f). A few small (5–7 μm) mono-phase, all-liquid fluid inclusions are present within this cement type.
C2: fine pore-filling equant sparitic calcite (crystal size <50 μm) with a constant bright orange–yellow cathodoluminescence response (Fig. 6d and h). No fluid inclusion has been found inside this cement type.
Scanning electron microscopy images showed the presence of clustered spherical bodies (1–2 μm in diameter) located around the Cm and Cr cements, which could represent bacteria. Details of the cement morphology and timing based on scanning electron microscopy photomicrographs is given in the Supplementary Material.
Cement types Cm, C1 and C2 are pore-filling and often prevent contact between quartz and feldspar clasts, giving rise to a floating fabric, which explains the high IGV. Tabular, elongate single, coalescent and nodular concretions show different amounts of these calcite cements. In particular, in elongate single and coalescent concretions, the pores are filled by almost equal proportions of C1 and C2 cements, whereas in the tabular bodies C1 becomes prevalent. Cr and Cm are rare and subordinate to C1 and C2 in all concretion types. Eventually, the cement within the nodular concretions is almost completely composed of C2 in a scattered pattern. Zoned C1 cement is subordinate and occurs as small crystals relative to those found in the tabular and coalescent concretions.
Stable carbon and oxygen isotopes
Bulk isotopic analyses (n = 184) on concretions from the Sabbie di Imola Formation show δ13C spanning from −10.7 to −2.5‰, whereas δ18O covers a narrower interval from −7.6 to −5.5‰ V-PDB (Fig. 7a). The isotopic data show a near-vertical alignment in a cross-plot (Fig. 7b). Nodular and small elongate single concretions have the least depleted δ13C values (−2.5‰), whereas tabular and elongate coalescent bodies usually show the most depleted isotopic response (−10.7‰). Data from a chimney-shaped concretion collected within the Argille Azzurre Formation for comparative purposes displays extremely depleted δ13C (−41.2 to −21.8‰) and slightly enriched δ18O (+2.8 to +5.1‰ V-PDB) values (Fig. 7a). Isotopic data obtained with the micromill (n = 45) allow a more precise evaluation of the isotopic signature of the main cement types (Fig. 7c–e). In single elongate concretions, the C2 cement has an isotopic signature similar to C1: δ13C ranges from −7.2 to −4.4‰, whereas δ18O spans from −8.5 to −5.9‰ V-PDB (Fig. 7c). Conversely, in the tabular and coalescent concretions, the C2 cement has less depleted δ13C (−6.4 to −4.5‰) than the C1 cement (−8.8 to −5.5‰), whereas the δ18O spans from −8.8 to −6.0‰ V-PDB (Fig. 7d and e). Statistical analysis to assess the difference in isotopic signature of the C1 and C2 cements can be found in the Supplementary Material. No sample from Cr and Cm could be isolated due to their size, which were below the sampling resolution of the micromill device.
Isotopic data collected on (a, b) hand specimens and (c–e) on thin sections with micromill technique. (a) Cumulative isotopic data characterizing the concretions inside the Sabbie di Imola and the Argille Azzurre formations. (b) Detailed isotopic trend shown by concretions hosted within the Sabbie di Imola Formation. Isotopic data gained from (c) single elongate, (d) coalescent elongate and (e) tabular concretions with micromill technique.
Along the isotopic sampling lines, single elongate concretions show slightly less depleted δ13C values near the centre along the horizontal transects, whereas δ18O remains almost constant. Both δ13C and δ18O are almost constant on vertical transects (Fig. 8a). Coalescent concretions display less depleted δ13C and more depleted δ18O values around the centre. Approaching the outer margin, δ13C shifts towards the more depleted values, whereas δ18O displays slightly less depleted signatures (Fig. 8b). Tabular concretions record oscillating δ13C and δ18O values along the horizontal transects. On vertical transects, moving away from the centre, δ13C becomes more depleted, whereas δ18O shows less depleted values (Fig. 8c). Nodular concretions are characterized by an alternation of less depleted and more depleted δ13C values and stable δ18O values of c. −6.0‰ V-PDB along the horizontal transect. On the vertical sampling lines, δ13C shows less depleted values in the central part, decreasing towards the outer margin. The δ18O values show only small oscillations around −6.5‰ V-PDB (Fig. 8d).
Sampling methods and isotopic trends on representative concretion hand specimens. (a) Single elongate, (b) elongate coalescent, (c) tabular and (d) nodular concretions.
Discussion
Carbonate concretion diagenetic history
According to the petrographic observations and stable isotope data, four cementation stages are proposed (Fig. 9a).
The first cement is Cr (Fig. 9b). The isopachous texture and dark cathodoluminescence pattern possibly suggest an oxidizing (high pO2) meteoric phreatic environment (Moore 1989; Budd & Land 1990; Hiatt & Pufhal 2014).
The second cement to precipitate is Cm (Fig. 9b). The dominant dark cathodoluminescence response suggests an environment with stable oxidizing geochemical conditions. The mosaic texture and the presence of meniscus cement around quartz and feldspar clasts support cementation in a meteoric vadose zone above the water table (Moore 1989) (Fig. 9a).
The third, more pervasive, cementation stage consists of C1 (Fig. 9b). The zonation is induced by changes in the water geochemistry (Habermann et al. 1996, 1998; Hiatt & Pufhal 2014), possibly caused by Eh and pH oscillations. It testifies to shifts from fully oxidizing to partially reducing to suboxic conditions and vice versa (Barnaby & Rimstidt 1989). This feature is common in meteoric phreatic environments with increasing reducing conditions moving away from the water table with depth (Li et al. 2017). The presence of zoned crystals could be related to cementation occurring within the meteoric phreatic zone in a deeper vertical position with respect to the Cr and Cm cements (Fig. 9a and b). The depleted δ13C and δ18O values of this cement support a meteoric origin of fluid with a contribution from soil-derived CO2 (Hudson 1977). The mono-phase fluid inclusions also support the hypothesis that C1 precipitated in the temperature range <40–50°C (Goldstein & Reynolds 1994).
The fourth cementing stage consists of C2 (Fig. 9b). The homogeneous bright orange cathodoluminescence response could be related to an environment with stable reducing conditions (low pO2) (Barnaby & Rimstidt 1989; Hiatt & Pufhal 2014). The δ13C is less depleted than C1, probably due to a lower contribution from soil-derived CO2 caused by the deeper conditions (Fig. 9a).
(a) Diagenetic environments inferred from the four identified cementation stages. (b) Schematic cementation sequence deduced from petrographic and cathodoluminescence observations. Cement colours refer to the cathodoluminescence response. Cr, isopachous rim of equant calcite; Cm, meniscus mosaic sparitic calcite; C1, blocky zoned calcite; C2, fine equant sparitic calcite; B, bioclast; Crd, crinoid; K-f, feldspar; Qz, quartz.
The reconstructed diagenetic sequence reflects the vertical variation of the diagenetic settings, with a progressive deepening trend (Fig. 9a and b). This could be due to the growth of the anticline southwards of the study site, causing a steepening of the hydraulic gradient or an increase in rainfall related to climate change during the diagenetic history.
The studied concretions developed in a shallow diagenetic setting, as confirmed by the limited thickness of the overlying strata, the presence of floating grains surrounded by cement and the high values of IGV (Table 1), which collectively indicate a maximum burial depth of c. 100–120 m (Houseknecht 1987; Cibin et al. 1993). The Sabbie di Imola Formation sand strata were deposited in a shallow marine setting along proximal fluvial lobes close to the shore. However, no evidence of marine cement has been found. This could mean that the sediments underwent a rapid change in sedimentary environment caused by a period of sea-level fall, in agreement with Gunderson et al. (2014), or uplift induced by the growth of the Quattro Castella Anticline (Ponza et al. 2010; Gunderson et al. 2014; Oppo et al. 2015). An alternative hypothesis is that the onset of topography-driven meteoric flow affected the permeable fluvial sand lobes after their deposition in a shallow marine setting (Chafetz et al. 1988).
Concretion growth mode
The size and arrangement of concretions depend on several factors, such as: (1) the spatial distribution of cementation nuclei; (2) the sedimentary facies; (3) the grain size–permeability structure; and (4) the fluid flow pattern.
Cementation nuclei are the sites where calcite precipitation starts (Bjorkum & Walderhaug 1990; Abdel-Wahab & McBride 2001). Within the studied concretions, the early cement generations (Cr and Cm in Fig. 6b and d) are often associated with fossil shells and detrital calcite grains, the limited dissolution of which probably influenced the fluid saturation conditions.
Several researchers have linked the presence of carbonate concretions to peculiar sedimentary facies (Cavazza et al. 2009; Van Den Bril & Swennen 2009; Arribas et al. 2012). Along the Enza River section, the occurrence of concretions is not strictly constrained to specific facies. However, the outcrop-scale three-dimensional facies distribution and geometry influenced the development of the concretions. At the lamina scale, the concretions cut lamina sets of ripple and hummocky cross-stratification with different angles, implying that the internal organization of facies did not exert a direct control on the development of concretions. This could be due to a low-permeability contrast between the different lamina sets or to fast advective flow forcing the fluids to neglect the internal structure of the facies.
Grain size and permeability distributions associated with the host sediments were the main drivers for the development of concretions (Mozley & Davis 1996; Davis et al. 2006). The studied concretions developed in sediments characterized by a mean grain size >90 μm and a permeability >700 mD. Concretions did not develop below these two threshold values, possibly because the limited fluid flow through the sediments was unable to provide the constituents needed for cementation.
Single elongate concretionary bodies represent the first stage of concretion growth (Fig. 10a and d). Such elongate bodies (Fig. 3a) preferentially develop parallel to the advective fluid flow direction (McBride et al. 1994; McBride & Parea 2001; Balsamo et al. 2012). Their occurrence suggests that, after the early cementation, the system was no longer saturated with respect to calcite locally and was unable to promote any further broadening of the concretionary bodies (Bjorkum & Walderhaug 1990). Cementation continued at other sites where the system was still saturated and closely spaced blade-shaped concretions merged together via newly formed cement bridges (Fig. 10b). Persistent and laterally extensive fluid flow eventually allowed the cementation of many elongate concretions to form tabular bodies parallel to the surfaces of the basal and upper strata (Fig. 10c). Conversely, elongate coalescent concretions originated from the focused advective fluid flow caused by the three-dimensional outcrop-scale facies distribution (Fig. 10d). Single elongate bodies progressively merged together (Fig. 10e) and eventually formed metre-wide concretions (Fig. 10f). Nodular concretions with a spherical shape are probably related to a steady, diffusion-dominated fluid flow regime (Theakstone 1981; Bjorkum & Walderhaug 1990; Wanas 2008).
Concretion development modality for (a–c) tabular and (d–f) coalescent elongate concretions. (a) Alignment of cementation nuclei leads to the formation of single elongate concretions in a pattern parallel to the strike of strata. (b) With a laterally spread fluid flow pattern (parallel to the bedding surface), single elements may merge together forming tabular bodies. (c) Pervasive cementation eventually allows the development of widely extensive tabular concretions. (d) The scattered and sparse arrangement of cementation nuclei is responsible for the formation of densely clustered arrays of single elongate concretions. (e) Focused fluid flow pattern induces the cementation of larger elongate bodies. (f) Coalescent elongate concretions increase their size until the fluid flow is able to carry the solutes necessary for cementation.
The proposed evolutionary pathways is consistent with the outward propagation of the cementation front (Bjorkum & Walderhaug 1990, 1993; Klein et al. 1999; Abdel-Wahab & McBride 2001; Mozley & Davis 2005), which agrees with the observed outward-depletion concentric pattern of δ13C values. The along-strike oscillatory pattern of δ13C values indicates the non-cylindrical, three-dimensionally complex outward propagation of cementation fronts and the possible interference among multiple growing cementation sites.
Sources of carbonate cement
The depleted δ13C values and the vertical alignment in the δ18O–δ13C cross-plot of cement isotopic data shown by concretions within the Sabbie di Imola Formation are consistent with a carbonate source mainly provided by the infiltration of CaCO3-saturated meteoric fluids carrying soil-derived CO2 (Fig. 7a and b) (Hudson 1977; Nelson & Smith 1996) facilitated by the shallow burial depth (<120 m). Further evidence of the meteoric origin of the fluids comes from the back-calculated δ18O composition of the cementing fluid. Assuming a superficial temperature of 12°C, a normal geothermal gradient (25°C km−1) and 120 m of maximum burial, the Friedman & O'Neil (1977) fractionation formula for the water–calcite system gives δ18O from −9.2 to −6.1‰ V-SMOW. These values are in accordance with the expected isotopic composition of meteoric water at the mountain front of the Northern Apennines (Giustini et al. 2016). Additional carbonate could be provided by the partial dissolution of detrital calcite and foram and mollusc shells (Bjorkum & Walderhaug 1990; Wanas 2008; Van Den Bril & Swennen 2009). The limited presence of bacterial cells inside fossil shells and the occurrence of framboidal pyrite suggest that bacterial activity may have influenced the distribution of nucleation sites during early cementation (Folk 1993; McBride et al. 1994; Abdel-Wahab & McBride 2001; McBride & Parea 2001).
Methane-rich fluids responsible for the formation of chimney concretions did not contribute to the cementation of the studied concretions inside the Sabbie di Imola Formation, as confirmed by the difference in the isotopic signatures (Fig. 7a).
Anticline growth and related fluid flow pathways
The growth-wedge geometry of the studied succession (Fig. 2b) and the clockwise rotation of the bedding dip direction indicate sedimentation during the growth of the anticline (Figs 4 and 5). In particular, during the deposition of the transition between the Argille Azzurre and the Sabbie di Imola formations, the strata were dipping towards the west. The carbonate concretions in this stratigraphic interval are aligned along the same westwards direction (Fig. 4b and c). The bedding dip direction shifted during the deposition of the Sabbie di Imola Formation from NW at the base to NNE at the top; the same trend is shown by the concretions (Fig. 4b and c). We therefore propose that the palaeo-groundwater flow recorded by the concretions was controlled by changes in the topographic gradient induced by the lateral propagation of the anticline (Fig. 11). This is also confirmed by the nearby streams changing flow direction from parallel to orthogonal with respect to the Apennines topographic front during fold amplification (Maestrelli et al. 2018). In particular, during the early lateral propagation stage of the anticline, the higher topography was positioned to the east of the studied outcrop (Fig. 11a), thus promoting a topographic and hydraulic gradient towards the west. After the incorporation of the studied outcrop within the fold forelimb, the topography increased rapidly in the crestal region of the fold (Fig. 11b), forcing groundwater fluxes towards the NW (Fig. 11c) and eventually to the NNE during further fold amplification (Fig. 11d). The evidence that the older concretions do not record the growth direction of the younger concretions implies that the geometry of the concretions only recorded the first fluid flow direction within the host sediments. This means that the first fluid flow direction imprinted the shape of the early concretions, whereas later groundwater flow probably increased the size of the concretions without affecting the original shape and elongation direction. The outer layer of the pre-cemented bodies acted as a nucleation surface (Bjorkum & Walderhaug 1990) for the new calcite precipitation provided by later fluid flow.
Conceptual sketch with inferred fluid flow patterns induced by anticline growth from the carbonate concretion data. (a) In the first stage of fold growth, the anticline was propagating near the study site and hence the fluid flow and sediment supply were directed parallel to the fold axis from east to west. (b) With an increase in topography southward from the Enza River, the fluid flow direction started to rotate clockwise towards the NW. (c) The westward propagation of the fold axis led to shallow fluxes towards the north. (d) During the late stage of fold growth, shallow fluid flow was mainly directed to the NNE following the hydraulic and topographic gradient induced by the growing anticline.
Conclusions
The studied concretions within the Pleistocene syn-kinematic sediments of the Sabbie di Imola Formation exposed along the Enza River stratigraphic section are interpreted to record different past patterns of groundwater fluid flow.
Four major conclusions can be drawn from our results.
Fossil fluid flow through weakly lithified sediments is still recognizable in the form of the selective cementation of tabular, elongate and nodular carbonate concretions and the flow direction can be inferred from the long axis of the concretions.
The carbonate concretions nucleated just after the deposition of sand in a transitional environment between meteoric vadose and meteoric phreatic. Cementation continued in a meteoric phreatic environment with the contribution of fluids percolating through soils. Concretions developed according to the concentric growth model, with advective flow forming elongate and tabular concretions, whereas nodular to spherical bodies grew under diffusion-dominated mass transfer conditions.
Concretionary bodies are selectively hosted in sediments with a permeability between 7 × 102 and 7 × 104 mD and a mean grain size from 90 to 290 μm, regardless of the internal structure of the sedimentary facies.
The clockwise rotation shown by the elongation direction of the concretions indicates that the lateral propagation of the Quattro Castella Anticline promoted a change in the topographic and hydraulic gradient along its western periclinal termination. The mean groundwater fluid flow direction locally changed from east–west (axial parallel) to south–north (axial orthogonal).
Acknowledgements
The reviewers R.T. Williams and G. Rawling and the Editor V. Vandeginste are warmly acknowledged for their suggestions and constructive criticisms, which helped to improve the final version of this paper. The authors thank E.M. Selmo and A. Di Matteo (University of Parma) for the stable isotope data. A. Comelli and L. Barchi (University of Parma) are thanked for thin section preparation and support during the scanning electron microscopy observations, respectively. S. Boullion, D. Vanhove and Z. Kelemen (KU Leuven) are kindly acknowledged for their support during micromill sampling. M. Pizzati collected field and laboratory data and wrote the paper; F. Balsamo participated in fieldwork and data interpretation; F. Storti conceived the research and took part in data interpretation; M. Mozafari and R. Swennen participated in the interpretation of diagenetic data; P. Iacumin contributed to the interpretation of the stable isotope data; R. Tinterri participated in the facies interpretation.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Scientific editing by Veerle Vandeginste
- © 2018 The Author(s). Published by The Geological Society of London. All rights reserved
References
From axial parallel to orthogonal groundwater flow during fold amplification: insights from carbonate concretion development during the growth of the Quattro Castella Anticline, Northern Apennines, Italy
