The onset of the sedimentary cycle in a mid-latitude upland environment: Weathering, pedogenesis, and geomorphic processes on plutonic rocks (Sila Massif, Calabria)
Published:January 01, 2007
- PDF LinkChapter PDF
Fabio Scarciglia, Emilia Le Pera, Salvatore Critelli, 2007. "The onset of the sedimentary cycle in a mid-latitude upland environment: Weathering, pedogenesis, and geomorphic processes on plutonic rocks (Sila Massif, Calabria)", Sedimentary Provenance and Petrogenesis: Perspectives from Petrography and Geochemistry, José Arribas, Mark J. Johnsson, Salvatore Critelli
Download citation file:
This work represents an integrated analysis of weathering landforms, including minor landform morphologies and soil profiles developed on granitoid terrains of the Sila Massif uplands (Calabria, southern Italy). The results of our analysis indicate that cryoclastic and thermoclastic processes, along with chemical weathering, are the main factors controlling rock degradation. Microscale features observed in primary minerals and parent rock fabrics, such as structural discontinuities, cleavage planes, fracturing patterns, and variations in chemical composition, play important roles in triggering weathering and, given sufficient time, progressively lead to grussification and soil development. Exfoliation, hydration, and splitting apart of biotite, as well as hydrolysis and etching of plagioclase and K-feldspar, appear to be prominent factors in the breakdown of bedrock.
Whereas time controls the degree of development of the main weathering features and climate infiuences type and intensity of the dominant processes, relief strongly influences the development and preservation/removal of the regolith/soil cover. Geomorphological evidence of severe surface erosion is quite good, especially along steep slopes where weathering products are quickly removed, although on the highest, dissected paleosurfaces (the oldest paleolandscape remnants in the Sila Massif), wide boulder fields represent relics of past, deep spheroidal weathering that have been exhumed by intense erosion. Erosive, depositional, or reworking phenomena, often enhanced by human activity, are well recorded by macro- and micromorphological features of soils, which show simple, poorly differentiated, rejuvenated profiles, buried or truncated horizons, abundant coarse-grained primary minerals or rock fragments, and pedorelicts. The soil clay mineralogy, characterized by illite, chlorite, and vermiculite, and the dominance of coarse textures confirm a young pedogenetic stage of evolution, although highly weathered sand grains (quartz included) occur in rarely preserved mature paleosols. This interpretation is also consistent with the compositional immaturity of fiuvial sands, which have undergone low to moderate transport.
Physical and chemical modifications of sediment prior to final deposition and burial reflect all aspects of the drainage area, including source rock texture and mineralogy, topography, climate, and transport processes (Basu, 1985; Johnsson, 1993; Le Pera et al., 2001; Arribas and Tortosa, 2003; Critelli et al., 2003). The contribution of various source-rock types to terrigenous sediments is largely dependent on the intensity of weathering, which may affect some rocks differently than others (e.g., Palomares and Arribas, 1993), and relief (Basu, 1985). For example, different types of bedrock may react differently to chemical weathering, resulting in variation in landscape and the development of weathering profiles (Ollier, 1971; Carson and Kirkby, 1972; Dixon and Young, 1981; Dejou et al., 1982; Pye, 1986; Twidale, 1990; Le Pera and Sorriso-Valvo, 2000b). Moreover, tectonics and climate, coupled with a combination of both chemical and physical weathering processes, may play a complementary role by acting as the main producers of sediment and regulating erosion rates (Le Pera and Sorriso-Valvo, 2000a; Riebe et al., 2000, 2001; Le Pera et al., 2001; Critelli and Le Pera, 2002; Scarciglia et al., 2005a). Furthermore, the genesis of clastic sediments and soils has been investigated in order to quantify processes occurring within source areas, including chemical and physical weathering, and textural and compositional modification of detritus during transition from bedrock to grus to soil to fluvial environment, in a well-controlled setting (e.g., Basu, 1985; Velbel, 1985; Grantham and Velbel, 1988; Cullers et al., 1988; Nesbitt and Young, 1989; Johnsson, 1993; Le Pera and Sorriso-Valvo, 2000b; Le Pera et al., 2001; Critelli and Le Pera, 2002; Girty et al., 2003). The Sila Massif (Fig. 1) is ideally suited to study relationships between landscape evolution and the genesis of clastic sediments and soils, and to calculate mass balances in a small, but tectonically active area characterized by very high sediment production (e.g., Ibbeken and Schleyer, 1991). In this work, we document modifications of source rocks through various stages of landform development within the Sila Massif and Neto River basin. Geologic (Critelli, 1999; Galli and Bosi, 2003) and morphologic (Critelli et al., 1991; Le Pera and Sorriso-Valvo, 2000a, 2000b; Le Pera et al., 2001; Molin et al., 2004; Scarciglia et al., 2005a, 2005b) evidence indicates that the study area is undergoing active landscape evolution.
GEOLOGICAL AND GEOMORPHOLOGICAL SETTING
The study area, located in the Sila Massif (Fig. 1), represents a section of the Hercynian orogenic belt of Western Europe, where allochthonous crystalline basement rocks during Miocene times were emplaced over Mesozoic to Cenozoic terrains of the southern Apennines. Rocks of the Sila Massif form the highest tectonic units (Calabrian arc) of the southern Italian fold-and-thrust belt (Amodio-Morelli et al., 1976) and consist of Paleozoic intrusive and metamorphic rocks, covered in places by unmetamorphosed Mesozoic sedimentary rocks (the Longobucco Group). Toward the Ionian Sea (to the east of the Sila Massif), a sedimentary succession of Miocene to Pleistocene age lies unconformably over the Paleozoic or Mesozoic rocks (Roda, 1964; Critelli, 1999). Paleozoic rocks consist of gneiss, amphibolite, schist, and phyllite, all of which have been affected by various Alpine metamorphic events and intruded by late Hercynian plutons (the Sila Batholith) (Messina et al., 1991). Gneiss consists of massive to migmatitic biotite-sillimanite-garnet–rich sills. A network of pegmatite dikes and an irregular thermal aureole of amphibolite facies rocks mark the contacts between Paleozoic country rocks and plutons (Caggianelli et al., 1994). The Sila Batholith consists of an array of intersecting intrusions, with different texture and fabric, that range in composition from granodiorite to gabbro and leucomonzogranite (Messina et al., 1991). Fission-track thermochronology of apatite and zircon from the basement rocks of the Sila Massif indicates a major period of exhumation from ca. 35 to ca. 15 Ma (Thomson, 1994).
The study area consists of flat to gently inclined paleosur faces, i.e., uplifted and dissected landforms formed during Pliocene(?)–Lower Pleistocene time (Dramis et al., 1990; Sorriso-Valvo, 1993; Matano and Di Nocera, 1999). These paleosurfaces occur at elevations between 1700 and 1000 m above sea level and developed across Paleozoic plutonic rocks and their Miocene-Pliocene sedimentary cover. Quaternary lacustrine and fluvial deposits, often terraced, are morphologically entrenched within the paleosurfaces of older landscapes (e.g., Sorriso-Valvo, 1993; Scarciglia et al., 2005a, 2005b).
The upper reaches of the main drainage systems of the Sila uplands, i.e., the Mucone, Savuto, Trionto, Nicà, and Neto Rivers, mainly drain plutonic rocks of the Sila Batholith, where weathering and soil profiles were described and sampled for this study. Within the intermediate and lower coastal reaches, these rivers drain metamorphic and/or sedimentary terranes. The Neto River, selected for quantitative compositional analyses, drains areas underlain by essentially Paleozoic granites and minor gneiss in its upper reaches, and Miocene to Pleistocene sedimentary rocks (siliciclastic and minor evaporites) in the lower reaches, where it feeds a modern delta system (Le Pera et al., 2001).
CLIMATE AND VEGETATION
The study area has a moist temperate climate, typical of upland Mediterranean zones (Csb, sensu Köppen, 1936), with warm, humid, and short summers and relatively mild winters. Mean monthly temperatures of the coldest month (January) are close to −1 °C/1 °C, whereas in July or August, they reach 16–18 °C. From November to April, daily temperatures commonly fall a few degrees below zero, with absolute minima reaching −10 °C or lower values; during summer, absolute maxima may attain 30–32 °C. Rainfall is spread throughout the winter season, with mean annual rates between 1000 and 1400–1600 mm. During the summer, rainfall never exceeds 30 mm (Versace et al., 1989; Lulli et al., 1992; Colacino et al., 1997). Snowfall occurs mainly in areas above 1400–1600 m, where it persists for ∼6 months (Lulli and Vecchio, 2000). According to the U.S. Department of Agriculture (2003), the pedoclimatic regime is mesic and udic (ARSSA, 2003).
Vegetation in the Sila uplands consists of grassland (often used for grazing), high mountain belt conifers (dominated by pine and fir) and/or beech forests, and cultivated fields. It is the result of a recent renaturalization and reforestation policy that was promoted during the last half-century after repeated earlier phases of severe deforestation, urban settlement, and poor agricultural practices: human activity had promoted progressively more extensive anthropogenic impact on the Sila Massif territories since the early prehistoric up to the classic (Greek and Roman) historical civilizations and modern times (e.g., Sorriso-Valvo, 1993; Scarciglia et al., 2005b).
Macroscale field studies, as well as micromorphological investigations of weathering and soil profiles developed on granitoid rocks, were carried out in a broad upland plateau area, and along different flat hilltops of some river catchments (Fig. 1).
Undisturbed samples were collected from diagnostic soil horizons, grus, and bedrock, and were impregnated with a polyester crystic resin and hardened for thin section preparation. Weathering and pedogenetic features were described and classified according to the scheme of FitzPatrick (1984). Scanning electron microscopy and energy dispersive spectroscopy (SEMEDS) were performed on sand-textured mineral grains and soil matrix clays. Samples were mounted on aluminum stubs, coated with gold, and examined with a Stereoscan 360 scanning electron microscope (Cambridge Instruments), equipped with an energy dispersive X-ray analyzer with a Si/Li-SUTW detector (EDAX, Philips Electronics). X-ray intensities were converted into weight percentages of oxides by applying the ZAF standard matrix correction procedure (Wilson, 1987); measured X-ray peaks were multiplied by factors dependent on atomic number (Z), absorption (A), and fluorescence (F) properties of specimens, to improve accuracy of chemical concentrations and mitigate errors. X-ray diffraction analysis (XRD) was performed on some selected soil horizons. Oriented specimens of the <0.2 μm clay fraction and random powders of the 2–0.2 mm sand fraction, which were separated by sieving and centrifugation following organic matter oxidation with H2O2, were scanned with a Philips 17/30 instrument (Cu-Kα radiation, 40 kW, 20 mA). X-ray diffractometer patterns were interpreted according to Berry (1974) and Brindley and Brown (1980). These approaches were utilized in order to recognize primary minerals and their pattern/degree of weathering, as well as pedogenetic clay minerals.
Finally, sand samples from the main channel of the Neto River were collected for point-count analysis (medium sand fraction, 0.50–0.25 mm).
The granitic rocks that crop out in the study area are characterized by extremely different degrees and patterns of physical and/or chemical weathering. The bedrock often shows a variety of physical discontinuities, such as shear zones, fault planes, pressure release and other types of jointing, pegmatite dikes, and minor lithologic or compositional changes. Such features sometimes intersect, creating blocks and wedges of different shapes and sizes. Overprinting this block and wedge pattern is an intense grussification, characterized by minute rock flawing and crumbling; this forms a gritty, loose material made of small polymineral aggregates and monomineralic particles (grain-by-grain disintegration, sensu Butzer , or arenization, sensu Power and Smith  and Teeuw et al. ). Very often, especially where the soil or grus cover is extremely thin or completely absent, tree roots are observed to penetrate the bedrock or saprolite at a depth of 1–2 m, divaricating and deepening preexisting joints.
Saprolite is commonly some tens meters thick, but may reach 100 m in depth (Cascini et al., 1992; Matano and Di Nocera, 1999). Rounded core-stones and boulders (Butzer, 1976; Ollier, 1988; Le Pera and Sorriso-Valvo, 2000a), ranging from ∼30 cm to 3–4 m in diameter (Fig. 2A, B), are often flaked, with an onion-like structure, and frequently occur within saprolite. They also may appear surrounded by a network of whitish to yellowish or light gray, bleached curved surfaces, intensely depleted in iron and fine particles (Fig. 2A). In contrast, other core-stones, and/or their grussified counterparts, often show reddish or blackish colors, related to Fe-Mn–oxide staining or clay illuviation. Similar features may coexist and form a mottled zone consisting of alternating subhorizontal or slightly wavy, whitish, eluvial surfaces, with yellowish brown, red, and/or black layers or elongated lenses. These features, even very close to the ground surface, may characterize the top of weathering profiles. In places, core-stones and boulders are partially (Fig. 2B) or completely exhumed (Figs. 2C and 2D) from the granular saprolite, and they frequently lie over the topographic surface, especially on the highest and oldest paleosurface remnants, and form wide boulder fields (Scarciglia et al., 2005a). Their unweathered inner core is usually surrounded by a weakly weathered, outer shell, with thin, millimeter- to centimeter-thick (Fig. 2D), fragmented exfoliation sheets (flaking; sensu Ollier, 1967, 1971) and encrusting lichens. The original translucent appearance of some primary feldspars in these external flakes has been replaced by a whitish to pale yellow alteration product with a dirty appearance, whereas biotite may exhibit cleaving and oxidation, coupled with local reddish Fe-staining as a possible biotite crystal derivative. On some boulders, massive exfoliation (large-scale spalling or sheeting; sensu Ollier, 1967, 1971) is present (Fig. 2C), with curved, edge-tapering sheets, 10–30 cm thick, as well as occasional splitting into large blocks along planar or curved cracks.
Saprolite and grus occasionally undergo coarse block rockfalls or trigger granular debris flows, mainly on the steepest slopes, where their stability thresholds are more easily overcome. Foot-slope scree taluses or detrital cones (Figs. 3A and 3B) are, in places, dominated by coarse rock fragments (gravels to boulders), while at other sites, they are mainly characterized by silt and sand to fine gravel and/or by a brownish to yellowish-brown or reddish (sometimes organic-rich) silty-clay matrix. Fallen tree trunks occur or lie over some scree talus accumulations (Fig. 3B).
The soil cover of the study area is not very homogeneous, and, in some zones, a complete lack of pedologic horizons allows the bare granite bedrock or saprolite to be exposed at the surface. Many soils are poorly differentiated and weakly structured and include a brown epipedon composed of an intimate mixture of humus and neoformed clay, which ranges from a few to some tens of centimeters in thickness (Fig. 4A). This organic-mineral horizon (A) is derived from the progressive degradation of fallen plant tissues and from the weak to moderate chemical weathering of primary minerals. Mainly, but not exclusively, on steep slopes, the tops of soils are often truncated or soils show abrupt lower boundaries with underlying bedrock. Soil profiles may reach a maximum depth of ∼1.5–2 m, where a flat topography and/or a protective forest cover occur, and/or along colluvial footslope belts, where they are affected by strong reworking and include frequent rock fragments or are buried by detritus and younger soils. At these sites, they appear better differentiated into various soil horizons that consist of simple cambic (Bw) and organic (O and A) horizons (Fig. 4B). Micromorphological observations of soils in thin sections permitted the identification of the following relevant pedogenetic features: (1) occasional rounded pedorelicts, i.e., reworked fragments of soils exhibiting features different from those characterizing the soils in which they are now included (Fig. 4C); (2) rare clay coatings and infillings, with slightly grainy to intensely speckled extinction patterns in crossed polarized light; (3) isolated fragments of clay coatings or infillings (papules; sensu Brewer, 1976); and (4) scarce to frequent silt or silty-clay coatings and cappings.
Phyllosilicates, such as illite, chlorite, and/or vermiculite, as well as some mixed layer components, dominate clay mineral assemblages. However, halloysite or possible short-range order aluminosilicate minerals are sometimes associated with these assemblages. More detailed results are available in Mirabella et al. (1996) and Scarciglia et al. (2005a, 2005b).
Bedrock and saprolite (R and Cr layers) are usually fragmented: very thin intragranular and transgranular microcracks may isolate small (millimetric to centimetric) mineral aggregates that are loosely bound together. Some feldspar grains in both bedrock and saprolite are quite weathered and have lost their original translucent aspect and acquired an almost powdery consistency, with dirty white to pale yellow colors. Micas, often affected by Fe-oxide staining, also exhibit distinct edge exfoliation or show an incipient disintegration by splitting along cleavage planes. Occasional to frequent Fe-depleted mottles and tongues, along with Fe-oxide segregations occur, as do rare to common clay coatings in intergranular voids.
Most of the soils are dominantly loamy sand to loam and sandy loam in texture. In addition, they display frequent skeletal grains (conversely having very low amounts of clay), umbric topsoils and cambic horizons, weakly to moderately acid pH, low cation exchange capacity, and small amounts of exchangeable bases, and an occasional lithic lower boundary toward the bedrock. In short, the main soil types are the orders Entisols and Inceptisols, sensu Soil Taxonomy (USDA, 2006); according to the WRB (World Reference Base for Soil Resources) classification (FAO et al., 1998), they are Leptosols, Cambisols, Umbrisols, Regosols, and Fluvisols (Dimase and Iovino, 1996; Lulli and Vecchio, 2000; ARSSA, 2003; Scarciglia et al., 2005a, 2005b).
More mature and older soils with other pedogenetic features, exceptionally reaching 5–6 m in depth, are exposed in extremely scattered and small sites (Dimase and Iovino, 1996; Lulli and Vecchio, 2000; ARSSA, 2003; Scarciglia et al., 2005a, 2005b). They mainly occur on terraced, granite-derived fluvial sediments, flat paleosurfaces, or colluvial slope deposits within protected landforms. Very often they may be truncated and/or buried by younger, less-developed soils, and they have a higher clay content and appear intensely rubified. Phyllosilicate clays include kaolinite (Fig. 5A) together with illite, and some chlorite and vermiculite minerals. Moreover, some are Alfisols (USDA, 2006) or Luvisols (FAO et al., 1998) with one or more argillic (Bt) horizon/s due to the abundant illuviation of fine material within pores and around soil aggregates. Microlaminated, crescentic clay coatings and infillings (Figs. 5B and 5C) were observed in thin sections. These pedofeatures frequently showed evidence of degeneration, with smooth-banded to speckled, grainy extinction patterns between crossed polars (Fig. 5B), evidence of fragmentation (Fig. 5C), and/or partial assimilation into the soil matrix (FitzPatrick, 1984).
The primary minerals that characterize the granitoids of the Sila uplands were identified in thin sections as well as from bulk samples (sand fraction) by means of optical microscopy (OM) and scanning electron microscopy with X-ray spectral analysis. For some samples, X-ray diffraction was also performed (Scarciglia et al., 2005a, 2005b), confirming the results from microscopy. The main components, both in bedrock and soil horizons, consist of quartz, K-feldspar, plagioclase, and mica (Fig. 6), and in soils, granite or granodiorite rock fragments. In particular, orthoclase and sometimes microcline occur, occasionally with Na-rich perthites, albite-twinned Na-plagioclase, biotite, and subordinate muscovite. In addition, accessory minerals, such as amphibole, apatite, epidote, Fe-oxide, and sphene, are present (Mirabella et al., 1996; Le Pera and Sorriso-Valvo, 2000a; Le Pera et al., 2001).
Many primary monocrystalline minerals and rock fragments, in the granite bedrock, grus, and soil horizons, show essentially mechanical weathering features, with irregular to linear or crossed fracture patterns. Regular fracture patterns follow main structural discontinuities, such as cleavage or twinning planes. Biotite crystals are frequently exfoliated and at least partly split into two or more fragments along cleavage planes (Figs. 6A and 6B). Small intragranular cracks often radiate from the cleaved biotite into surrounding feldspar and quartz crystals (Fig. 6A) and/or extend into intergranular cracks separating irregular polycrystalline rock fragments and monomineralic grains.
Weak to moderate chemical weathering affects crystals, more intensely in soil horizons than in grus layers and bedrock. Biotite may be oxidized, etched along its outer edges, or show evidence of early clay neogenesis (chloritization), as indicated by a local increase of Al and Mg in its chemical composition. K-feldspar and, above all, plagioclase frequently exhibit surface etch pitting, solution lines, Fe-oxide staining (Figs. 6C and 6D), and/or clay neogenesis. Na-perthite included within the K-feldspar component appears to be more highly weathered than the surrounding crystal (Fig. 6E). Occasionally, extremely weathered, unrecognizable pseudomorphs are found (Fig. 6F). More rarely, some quartz crystals in humic horizons show extremely superficial etch pits with subspherical, rounded shapes.
In the cambic or argillic horizons of the rare well-developed soils, e.g., those formed on the Middle to Upper Pleistocene terraced fluvial deposits near Cecita Lake (Scarciglia et al., 2005a, 2005b), quartz grains appear severely weathered. For example, some of them display physical fractures and signs of abrasion, while others exhibit microtextures that consist of irregular (Fig. 6G) or well-rounded to hexagonal, rhomboidal, and regularly oriented triangular etch pits, scarce linear, parallel-oriented grooves, and amorphous silica coatings or quartz overgrowths. Sometimes, solution features are particularly deep and may create an exceptionally vacuolar, cavernous structure in quartz grains (Fig. 6H).
FLUVIAL SAND, SOIL, GRUS, AND PARENT ROCK COMPOSITIONS
The detrital modes of the sand samples from the Neto River were compared to sands derived from grus and soil environments in the upper reaches of the same basin. Petrographic indices, such as the ratio of monocrystalline quartz (Qm) to total feldspar (F), and plagioclase (P) to total feldspar (F) ratio (P/F), indicate a nearly constant sand composition along the drainage basin of the Neto River.
Modal analysis of “grus-soil-fluvial sand” indicates that an increase in monocrystalline quartz grains (QmFLt) (Fig. 7) along with a decrease in plagioclase grains (QmKP) (Fig. 8) characterize the transition from bedrock to grus. Modal compositions of related grus and soil are characterized by an increase in monocrystalline quartz, coupled with a decrease in K-feldspar and plagioclase (QmFLt and QmKP diagrams, respectively) (Figs. 7 and 8). This trend has led to a mineralogic zonation of the soil profile (e.g., Nesbitt et al., 1997).
The Qm:F:Lt ratios in sands from the headwaters to the mouth of the Neto River are the same, although there are multiple source rocks throughout the basin (Le Pera et al., 2001). The composition of fluvial sand is nearly homogeneous and quartzofeldspathic (51%–48%–1% QmFLt in upstream sand; 53%–45%–2% QmFLt in downstream sand; 46%–45%–9% QmFLt in coastal sand); hence, there is no evidence that chemical weathering, sorting, and/or abrasion have affected feldspars relative to quartz in the fluvial system. Virtually all changes to the feldspar/ quartz ratio must have taken place within weathering profiles mantling the bedrock of the source area and before incorporation into the Neto River. These results agree with findings from various other river systems (e.g., Nesbitt et al., 1996; Franzinelli and Potter, 1983), even if a downstream increase in compositional maturity cannot be ruled out (Johnsson et al., 1991; Robinson and Johnsson, 1997).
The great variety of physical discontinuities affecting the granite bedrock, such as fault planes and shear zones, pressure release features, and lithologic changes, developed as a consequence of the complex tectonic history during the emplacement and uplift of the Sila Batholith. They all act as preferential weakness sites for the onset of both physical and chemical weathering processes.
Subspherical boulders and core-stones represent the inner, unweathered, or less-weathered portions of saprolite due to deep spheroidal weathering (e.g., Ollier, 1967; Twidale, 1986; Thomas, 1994; Migoń and Lidmar-Bergström, 2001), and they derive from progressive smoothing of the outer edges of jointed crystalline rocks to subspherical shapes, with complete isolation of rounded cores within a saprolitic to grussified groundmass. Spheroidal weathering appears to be the result of water percolation through rock fractures, a process that in turn enhances chemical reactions (hydration, solution, and mainly hydrolysis and oxidation) at the rock-water interface and thus widens and rounds off the sharp jagged edges of preexisting joints.
The discolored bleached zones surrounding core-stones and boulders within the saprolitic to grussified groundmass represent preferential pathways for downward-flowing water. Along such pathways, leaching of various components (fine particles or ions) and/or gleying are common processes. In contrast, the darker, reddish or blackish zones represent illuvial zones enriched in Fe- or Mn-oxides. The alternation of whitish and reddish/yellowish mottles in wavy, subhorizontal patterns indicates the contemporary occurrence of eluvial and illuvial zones, and/or it suggests alternating reducing and oxidizing conditions, possibly brought on by an oscillating water table. Where these features crop out near the ground surface, no evidence of a present-day water table is generally found. It can therefore be supposed that they represent signs of deep weathering, related to a paleohydrological (and presumably paleomorphological) feature, and that very intense erosion occurred, which removed the surface counterpart of such a feature (grus + soil cover). This interpretation is consistent with the occurrence of isolated spheroidal core-stones and boulders close to or at the ground surface of the oldest landforms in the Sila Massif, suggesting that they may have previously formed at considerable depths under constant volume conditions (Ollier, 1967, 1971) and were only later exhumed by erosion. At present, they denote an environment dominated by weathering-limited conditions, where transport rates of loose and mobile detritus strongly overcome those at which detritus is generated by weathering processes. The exfoliation of thick concentric shells (onionlike structure) observed on boulders is a typical feature of deep spheroidal weathering that occurs below the ground surface, where hydrolysis of silicate minerals dominates (Ollier, 1988). The efficacy of deep weathering in the study area can be related to high local relief, resulting from the strong tectonic uplift of the Sila Massif, and the concomitant high hydraulic gradient. The combination of these two features increases the deep penetration of groundwater for chemical reactions and the efficiency of frost weathering (see following). Once exposed at the surface, the weathering trend and modes presumably changed, involving volume increase, since boulders became free to expand. Boulder sheeting into thick slabs very likely represents an unloading feature (Ollier, 1967, 1971), whereas the large-scale splitting of some boulders along planar or curved cracks can be interpreted as a result of the induced tensile fracture mechanism proposed by Ollier (1978). Flaking, the surface type of exfoliation observed on these boulders, characterized by a small thickness of curved flakes, appears quite different from the spheroidal thick sheeting, and it is conceivably due to subaerial weathering (Ollier, 1967). It is very likely that weathering processes in the exposed environment preferentially act also on inherited (stress release and/or subsurface chemical weathering) features, possibly accelerating and amplifying their development.
Under the present-day climatic conditions, cryoclastic phenomena due to freeze-thaw cycles appear to be among the most effective physical factors for rock degradation during the coldest months, mainly widening and deepening joints and microcracks. Crack systems increase access of water to rock surfaces, accelerating the weathering processes and enhancing rock disintegration by frost weathering (Migoń and Thomas, 2002). Walder and Hallet (1985) demonstrated that crack growth is mostly effective if temperature reaches values between −4 and −15 °C, an uncommon temperature range in the study area. However, more recently Matsuoka (2001b) showed that at least partial filling of joints with water can minimize ice extrusion to open surfaces, thus concentrating volumetric expansion forces toward joint walls and enhancing the effectiveness of frost wedging even at not-so-low temperature values. This process is enhanced if unconfined water conditions are present (i.e., if water availability and migration through the pore-microfracture system occurs), even though oscillations around 0 °C do not take place (Matsuoka, 2001a). Although no data are available about amount and frequency of diurnal freeze-thaw cycles in the study area, oscillations a few degrees around zero can be presumed to occur rather frequently on the basis of the measured climatic parameters, along with the prolonged preservation of the snow cover during winter (especially on shadowed slopes or in protected patches). Such a conclusion is also supported by the occurrence of various cryonival morphologies and processes (Scarciglia et al., 2005b), for example, soil-grass hummocks (thufurs), shallow solifluction terraces, the growth of ice needles in topsoil horizons, the presence of illuvial silt pedofeatures, and the recurrent formation of night hoarfrost and its persistence for many hours after sunrise.
In addition, it can be assumed that cryoclastism was more efficient during the cold stadial phases of Quaternary glacial periods. The occurrence of poorly weathered but friable or disintegrated rock in the plateaus of the Sila Massif may potentially reflect cryogenic grussification (Butzer, 1976; Migoń and Thomas, 2002), which is at least partly inherited.
Similarly, thermoclastic processes likely play an important role for rock breakdown in the study area. According to Hall (1997) and Hall and André (2003), thermal stress fatigue and shock processes can cause rock shattering without the aid of freezing water, if very fast and frequent, low-wavelength, and low-amplitude thermal oscillations occur, possibly induced by wind refrigerant power. Also, diurnal to seasonal and yearly cycles of insolation can be assumed to be relevant. For example, different mineral species comprising the granitic bedrock have differential volumetric expansion/contraction responses to thermal variations, due to the different power of reflection-absorption-release of solar radiation by dark- and light-colored minerals (e.g., biotite and feldspar) (Ehlen, 2002). This results in different times and rates of grain microcracking. Moreover, insolation heating, which is thought to be particularly intense in upland areas, induces faster and larger temperature changes on the exposed rock surface than in its interior, thus producing wide temperature differences between the outer and inner parts. This implies that rapid temperature changes may lead to the continuous, differential expansion and contraction of rock minerals (Zhu et al., 2003), potentially creating thermal rock fatigue and breakdown. In addition, differential thermal expansion and displacement along different crystallographic axes typical of different mineral species may induce differential stress accumulation and release along preferential (structural) zones of weakness, also propagating into and destabilizing the surrounding rock mass.
These physical processes seem to be of prominent importance for the breakdown of granite, in particular, the small-scale exfoliation around boulders and the formation of grus. Furthermore, the effects of lichens (frequently observed on boulder surfaces) should not be neglected. It is well documented that they promote physical widening of fissures along their outer edges by adhering with their thallus and protruding their hyphae into the rock mass, and activate chemical dissolution by the production of acidic substances such as oxalic acids (Adamo and Violante, 2000).
Tree root penetration also plays a key role in rock wedging and fragmentation, as suggested by roots penetrating at depth within the cracked granite rock or saprolite, and occasional wind throw. The sporadic landslides triggered by rock degradation and breakage induce further fragmentation and comminution of rock particles. In particular, coarse debris accumulation on scree taluses and detrital cones is the result of sediment transport as rock falls and topples. Conversely, loose fine detritus (sometimes with higher clay and organic matter content) points to debris flow emplacement of intensely grussified rock and pedogenized material, probably enhanced by a prolonged period of weathering and pedogenesis, maybe under a stable forest cover (which is supported by the presence of fallen trees within the debris). All these processes produce sediment that enters the fluvial system.
Crystal microtextures are prominent in the weathering, fragmentation, and granular disintegration of the Sila Massif granitoids—they have an important upscaling effect—and in triggering or accelerating soil development. The main physical weathering features that affect mineral grains appear to be strictly controlled by major structural discontinuities (cleavage and twinning planes), or sometimes by irregular breakage patterns caused by sedimentary processes. A significant role in rock shattering seems to be played by biotite. Exfoliation and oxidation enhanced by hydration, coupled with progressive vermiculitization or chloritization due to hydrolysis, may lead to expansion and splitting apart of individual biotite flakes along cleavage planes (Taboada and García, 1999a, 1999b). This process creates a network of radiating intragranular and transgranular fractures that extend toward surrounding crystals (Isherwood and Street, 1976), which induce tension and consequent rock fatigue and breakdown. The observation that the amount of biotite crystals in soil sands derived from granites is higher than in correlated unweathered parent rock (Cullers, 1988; Le Pera et al., 2001) supports this hypothesis.
The principal chemical weathering features (etch pits or solution lines) and products of such processes (Fe-oxides and neoformed clay minerals) have been observed on feldspars (with plagioclase usually more weathered than K-feldspar) and micas. Etch pits and solution lines seem to be controlled by physical discontinuities (both natural and mechanically induced), coupled with possible differences in chemical composition (FitzPatrick, 1986), e.g., sodium microperthites within K-feldspar grains. Intragranular discontinuities become preferential sites for water migration and interaction with mineral surfaces, and, as a result, they enhance chemical attack. This attack results in the widening or deepening of etch pits and solution lines, with a progressive increase of available surface area, which in turn favors further chemical reactions and the weakening and disintegration of the whole rock mass. Etch forms grow and may coalesce eventually resulting in a complete pseudomorphic transformation. This general process is consistent with natural and experimental observations on alkali feldspars (Lee et al., 1998), which have shown a good correlation between dissolution rates and the density of micro-discontinuities, and they have shown that microtextures produce the greatest impact on mineral weathering rates during advanced stages of dissolution, when grains start to disintegrate to a microgranular material. The formation of secondary clay minerals along some microdiscontinuities could represent another source of fragmentation if they undergo expansion in the presence of water (cf. Frazier and Graham, 2000). The instability generated along intermineral contacts appears strongly favored in fine-grained rocks, which are characterized by a greater surface energy (Taboada and García, 1999a). Similarly, the occurrence of more intensely weathered mineral grains within soil horizons rather than in saprolite or parent rock suggests that, as soon as weathering processes develop, there is an increase in pedogenetic matrix produced by clay neogenesis and Fe oxidation. This increase in matrix, in turn, favors chemical reactions promoted by a prolonged interaction between mineral surfaces and the circulating soil solution. In addition, the activity of humic acids characteristic of organic-mineral horizons appears to enhance etching phenomena on primary minerals.
Some quartz grains from the most mature soils show cracks and signs of abrasion, probably related to crystal ruptures and impacts during sedimentary processes (Krinsley and Doornkamp, 1973; Al-Saleh and Khalaf, 1982) and in accord with their occurrence in soil profiles developed on fluvial deposits. The chemical microtextures on quartz grains display irregular to very regular outlines and distribution. Those grains that exhibit regular patterns clearly indicate a selective, crystallographic control over dissolution (Eswaran and Stoops, 1979; Al-Saleh and Khalaf, 1982; Howard et al., 1996). Some irregular solution features presumably represent the result of coalescing and deepening of previous smaller holes. The combination of secondary amorphous silica precipitates or quartz overgrowths (Krinsley and Doornkamp, 1973; Mazzullo and Magenheimer, 1987; Newsome and Ladd, 1999) with the quartz solution features suggests that they are conceivably the complementary facets of one main process.
Without excluding the well-assessed role of hydrothermal diagenetic processes to explain quartz dissolution (Dove and Nix, 1997; Dove, 1999), the severely etched quartz crystals can be interpreted to be the result of one or more of the following: (1) a highly acidic soil reaction initiated by humic compounds from organic-rich horizons (Krinsley and Doornkamp, 1973; Howard et al., 1996); (2) an aggressive pedoclimatic regime, under particularly warm and humid paleoclimates (e.g., tropical/subtropical environments), which enhanced chemical reactions and intense leaching phenomena and the ability to induce acidic pH values and silica solution (cf. Krinsley and Doornkamp, 1973; Eswaran and Stoops, 1979; Stoops, 1989; Summerfield, 1991; Pell and Chivas, 1995; Malengreau and Sposito, 1997; Moral Cardona et al., 1997); (3) a locally alkaline soil environment at the level of weathering microsites, with availability of bases (alkali and alkaline-earth cations) that promoted quartz solubility (Dove and Nix, 1997; Karlsson et al., 2001); and (4) a very long period of pedogenesis (Stoops, 1989; Al-Saleh and Khalaf, 1982; Howard et al., 1996; Schulz and White, 1999), which allowed polycyclic or polygenetic processes to occur.
Despite the fact that soils of the study area have subacid pH values (which favor only weak silica solution) and that only rarely do etched quartz grains occur in organic horizons, the role played by an acidic pedoenvironment can be of key importance for the formation of quartz chemical microtextures. For example, it is responsible for the hydrolysis of primary minerals, such as feldspars and micas, and thus increases the availability of bases (Scarciglia et al., 2005a); H+ ions are incorporated into the original crystalline lattice as base elements are leached out. The leaching of bases in turn enhances an overall acidic soil environment, which produces additional hydrolysis. A concentration of base elements may induce a locally alkaline reaction at the soil solution–mineral surface, a result that would strongly increase silica solubility. Past climatic conditions warmer and more humid than today (presumably past Quaternary interglacials) could have been particularly favorable to the development of the quartz solution features at issue, possibly also increasing silica undersaturation of the soil solution by intense leaching, which would have in turn induced a thermodynamic disequilibrium of quartz (Schulz and White, 1999). The presence of etched quartz grains in the older, mature soils is consistent with other features of these soils, discussed in the following paragraphs. Alternatively, time appears to be a major cause of the severe etching of quartz. Even quite small, highly localized initial variations in weathering resistance (fractures, microtopographic features, biotic impacts) may operate as instability factors for the onset of etching, so that minor, apparently imperceptible solution features are cumulative and amplified through time (Howard et al., 1996; Turkington and Phillips, 2004). The extremely cavernous aspect of some quartz crystals could be an effect of such an amplification, possibly due to deepening, widening, and coalescence of initial microtextures. The size of etch features, weathering rates, and moisture concentration appear strictly correlated in a self-reinforcing mechanism. For example, the weathering of cavity inner walls tends to enlarge the hollows, in turn concentrating moisture in the interior, resulting in a further increase in weathering rates. Hence, a climatically controlled increase in moisture supply would have accelerated weathering rates and improved cavern growth.
Intensely etched quartz grains occur in mature or ancient soils, as some extremely weathered feldspar and plagioclase grains do. The latter minerals occur with fresh or weakly altered feldspars in some young soils. Significantly, both of these occurrences are found in soils that developed on fluvial deposits. This relationship suggests a cumulative effect of long-term processes that act only on selected mineral grains as a consequence of the multicycle origin of their parent material. Some degree of weathering of primary minerals conceivably was inherited from previous pedogenetic cycle(s), which would clearly enhance more advanced in situ alteration and pedogenesis. The occurrence of pedorelicts supports the hypothesis of recycling (see following).
The coexistence of variable weathering patterns (macrocracking, granular disintegration, spheroidal boulder/core-stone separation, mottling and bleaching, etc.), some of which occur in single features, sometimes as assemblages of more than one, and different degrees of weathering spatially very close to one another in the study area suggest a prominent microclimatic control. Even small differences in insolation and moisture availability, possibly infiuenced by small lithologic, relief (elevation, slope, and aspect), and vegetation changes, can be supposed to affect the dominant type(s), rate(s), and pattern(s) of weathering processes. In addition, it is also very likely that certain features related to present-day weathering processes are possibly superimposed or associated with others inherited from relict environments and processes.
The dominance of simple, poorly differentiated profiles, which consist of organic (O and A) and possibly cambic (Bw) horizons and overlie the parent rock or sediments (R or C), indicates a poor to moderate pedogenetic maturity. This weak development is consistent with coarse-grained textures, which still preserve abundant primary minerals or rock fragments, and therefore a strong influence by parent rock. Also, the low amount of clay (and associated low cation exchange capacity values) can be interpreted as resulting from poor pedogenetic development of clay minerals derived from an overall weak degree of weathering of primary components. Further support for this conclusion is provided by soil clay mineralogy. Illite, chlorite, vermiculite, and halloysite may represent weathering products of feldspars and micas (Barnhisel and Bertsch, 1989; Blum and Erel, 1997; Kretzschmar et al., 1997; Taboada and García, 1999b; Thomas et al., 1999; Sequeira Braga et al., 2002). Hydration and/or isomorphous substitution of cations in the structural units of primary micas commonly leads to the formation of illite, chlorite, and vermiculite in the early stages of weathering, and they are frequently found in granitic saprolites in temperate climates (e.g., Sequeira Braga et al., 2002). All these features, coupled with the occurrence of buried or truncated horizons, abrupt boundaries between different soil horizons or between soil horizons and the bedrock, rounded pedorelicts and papules, suggest a complex history of erosion, deposition, or reworking. These processes induce and enhance soil rejuvenation on steeper landforms and through human activity (forest clearance, tillage, and pasture) produce mobile material that in turn migrates and enters the sedimentary cycle within the drainage system, or stops and persists for a certain time span, and is recycled and undergoes further in situ weathering and/or pedogenesis.
As a whole, soils of the Sila plateaus formed in a highly leached pedoenvironment, as indicated by low base saturation, small amounts of exchangeable bases, and acidic soil reactions. These features are in accordance with the perhumid climate of the study area, where prolonged moisture availability and well-drained conditions result in an udic soil regime. This conclusion is consistent with the leaching factor (sensu Crowther, 1930) of the Sila uplands, which, based on primary climatic parameters, is the highest in Europe (Le Pera and Sorriso-Valvo, 2000a).
A higher degree of pedogenetic evolution is recorded by less widespread soils that have reddish colors and higher clay content as a consequence of an advanced weathering of primary minerals. As far as chemical reactions (hydration, solution, hydrolysis, and reduction/oxidation) occur, the release of iron from primary components and the subsequent crystallization of Fe-oxides and hydroxides in the soil matrix, as well as the genesis of secondary clay minerals, tend to increase. Sometimes the increase in clay fraction is due to illuviation, evidenced by abundant laminated clay coatings. In particular, these pedofeatures are relict, as they are frequently fragmented, degenerated, and partially assimilated into the soil matrix (FitzPatrick, 1984; Catt, 1989; Kemp, 1998; Scarciglia et al., 2003a, 2003b, 2005b). Both rubification and clay illuviation require warm and humid climates with xeric regimes (sensu USDA, 2006). Abundant rainfall and consequent soil moisture availability trigger chemical attack of primary minerals (and possible Fe release from Fe-bearing components), as well as downward migration of clays suspended in the soil solution. A marked seasonal contrast produces a water deficit under free drainage conditions in the dry season, leading to the oxidation of iron and the reddening of the matrix on one hand (cf. Diaz and Torrent, 1989; Schwertmann and Taylor, 1989), and the adhesion of clay coatings in pores on the other (Fedoroff, 1997). Such conditions appear in contrast with the present-day pedoenvironment, and in accordance with the relict significance, they are ascribed to the degenerated clay coatings; warmer and more humid climates than the present possibly occurred during past Quaternary interglacials (Scarciglia et al., 2003a, 2003b, 2005a, 2005b). Moreover, the presence of pedogenic kaolinite suggests a humid soil environment, where hydrolysis of primary minerals was relatively intense, and well-drained conditions, which promoted leaching of silica and soluble cations; this favored the formation of 1:1 clay minerals (Summerfield, 1991; Righi et al., 1999), possibly associated with a long period of pedogenesis (Bronger and Bruhn, 1989). Kaolinite could have been derived from the alteration of biotite (Kretzschmar et al., 1997; Sequeira Braga et al., 2002), possibly through an intermediate product such as vermiculite, as documented in highly leached soils and saprolites in tropical regions (Blum and Erel, 1997). These hypotheses are in good agreement with the quartz solution features described earlier in this paper. As a whole, the main features observed in the scattered remnants of mature and older soils imply a higher land surface stability, favored by relatively flat topography, where retention of movable grus and soil material was favored and rates of weathering and pedogenesis exceeded rates of erosion. Moreover, the longer the time span of geomorphologic stability (and pedogenesis), the higher was the possibility that different climatic conditions alternated during pedogenetic cycles, which therefore also recorded possibly polygenetic or polycyclic processes.
Fluvial Sand Composition as an Indication of Weathering Zone Erosion
The composition of sand from the Neto River is the result of chemical weathering and erosion of source rocks and of negligible transport within the riverine system (e.g., Le Pera et al., 2001).
Comparison of the detrital modes of fluvial and weathering profile sand emphasizes the petrogenesis of sediments in the absence of prolonged transport processes (e.g., Cullers et al., 1988; Critelli and Le Pera, 2002). Modal data from the Neto River sand samples, and especially the high F/Q ratio, suggest that the major source for most of the sand is grus developed on granodiorite and monzogranite of the Sila Batholith.
Sand composition is quartzofeldspathic and nearly homogeneous along the main channel of the Neto River, even where it cuts across a blanket of sedimentary cover. Thus, fluvial transport processes do not alter sand composition within the Neto drainage basin, and the nearly constant sand composition indicates minimal sediment maturation. These feldspar-rich sands indicate erosion mainly from grus rather than soil horizons of the weathering profile (e.g., Nesbitt et al., 1997). Moreover, the short and rapid sediment transport by the river, along with very short sediment storage, could have inhibited the effects of weathering, leading to remarkably immature fluvial sands. High-gradient slopes within the river drainage system, as assessed in another river draining the Sila uplands (Le Pera and Sorriso-Valvo, 2000b) and the severe mechanical erosion of source rocks are conditions that result in production of immature sediment (Nesbitt and Markovics, 1997; Nesbitt et al., 1997). An increase in monocrystalline quartz and mica grains and a decrease of plagioclase during the conversion of granite bedrock to regolith could suggest a rapid loss of plagioclase during grussification. The enrichment of biotite crystals from bedrock to grus to soil horizons is consistent with micromorphological observations, which show the expansion and splitting of weathered biotite, preferentially along cleavage planes, which lead to a greater amount of biotite flakes (cf. Cullers, 1988; Le Pera et al., 2001). The enrichment of quartz and the depletion in K-feldspar and plagioclase from grus to soil microenvironments indicate a further intensification of alteration processes acting on labile minerals. This result fits within the realm of modifications taking place in the pedogenetic environment, which include biological activity (e.g., Basu, 1981; Moulton and Berner, 1998), a prolonged interaction of the soil solution with sandy detritus, and strongly enhance chemical attack of primary minerals. In fact, abrasion during bedload transport of the Neto River appears to be insufficient to initiate the comminution of K-feldspar and plagioclase grains (Le Pera et al., 2001); hence, these minerals are preferentially destroyed in weathering and soil profiles (Nesbitt et al., 1996, 1997).
This work shows the great potential of combining complementary scientific disciplines to the study of weathering profiles and the derivative sand. Both physical and chemical weathering affect plutonic rocks of the Sila upland, where tectonics plays a key role as a predisposing source of (1) rock alteration or fragmentation through discontinuities and strain features; (2) local relief and slope produced by uplift, which act as driving forces of the main morphodynamic processes (erosion, transport, deposition), on one hand, reducing the time that detritus spends in the soil profile and concomitant rates of chemical weathering, and on the other, exposing deep unweathered bedrock to the surface, thus rejuvenating the weathering front and continuously producing newly weathered loose material; and (3) an enhanced, high hydraulic gradient that is able to promote a deeper percolation of water and consequent penetration of weathering at depth. Hydrolysis and oxidation were identified as important chemical processes operative in weathering profiles, whereas frost shattering and thermoclastism are the dominant processes of physical rock breakdown. These phenomena are enhanced by microtextures on mineral grains and macroscale discontinuities, which mutually interact to allow water penetration and access to rock surfaces and progressively lead to rock disintegration. Hydration, expansion, and cleaving of biotite and feldspar (sometimes even quartz) play important roles in grussification and soil formation. Although the present-day humid climate of the Sila Massif is very favorable to weathering processes, modal compositions of fluvial sand samples from the Neto River, and the weak soil development, indicate low pedogenetic and compositional maturity of sediments and soils in the studied drainage basins resulting from soil rejuvenation by erosion and rapid fluvial transport. In contrast, ancient and mature soils, rarely preserved on some flat paleosurfaces, fluvial terraces, or colluvial belts, clearly denote a higher degree of weathering, possibly due to a local multicycle origin of the soil parent material.
As a whole, a weathering-limited erosion regime characterizes the Sila uplands, where steep slopes and high local relief promote short residence times of weathered loose material, and potential sediment transport rates tend to exceed those at which weathering mantles are produced. In contrast, transport-limited conditions prevail on geomorphologically stable landforms (mainly terraced surfaces and depressions), where sediment storage and weathering and pedogenesis rates clearly overcome the efficiency of regolith and soil removal. Results discussed herein demonstrate that the composition of fluvial sand derived from weathering and soil profiles in the Neto basin is strongly dependent upon the extent of in situ chemical weathering of source rock. The composition of sands highlights that sediment of the Neto River is mainly derived from grus: the fluvial sands are composed dominantly of quartz and feldspar, which mimic that of sand-grade material studied from the grus. We conclude that first-cycle sands of the Neto River do not reflect the unweathered plutonic bedrock but the weathering profile environment, and specifically, the zone in which grus forms, widely mantling it. In this regard, our study provides confirmation (Grantham and Velbel, 1988; Suttner et al., 1981; Nesbitt et al., 1997; Le Pera et al., 2001; Girty et al., 2003) that chemical weathering and pedogenesis represent a fundamentally important control on the petrogenesis of siliciclastic sediments rather than provenance. Both climate and duration of weathering and pedogenesis appear to have a prominent influence on weathering rates and intensity.
We are grateful to Mark J. Johnsson, Robert Cullers, and Gary H. Girty for their critical comments and suggestions, with special gratitude to the latter reviewer for his patience and editing care, which improved the quality of the manuscript. Many thanks are also due to Kathie Marsaglia for her fruitful revision of the final version of the manuscript.
Figures & Tables
Sedimentary Provenance and Petrogenesis: Perspectives from Petrography and Geochemistry
- Calabria Italy
- chemical weathering
- clastic sediments
- clay mineralogy
- climate effects
- human activity
- igneous rocks
- Mediterranean region
- mineral composition
- parent materials
- plutonic rocks
- sedimentary rocks
- SEM data
- Sila Massif
- soil profiles
- Southern Europe
- southern Italy