Paleopedology as a Tool for Reconstructing Paleoenvironments and Paleoecology

Abstract

Soils form as a product of physical, chemical, and biological activity at the outermost veneer of Earth’s surface. Once buried and incorporated into the sedimentary record, these soils, now paleosols, preserve archives of ancient climates, ecosystems, and sedimentary systems. Paleopedology , the study of paleosols, includes qualitative interpretation of physical characteristics and quantitative analysis of geochemical and mineralogical assays. In this chapter, the paleosol macroscopic, micromorphological, mineralogical, and geochemical indicators of paleoecology are discussed with emphasis on basic analytical and interpretative techniques. These data can reveal a breadth of site-specific interpretations of vegetation, sedimentary processes, climatic variables, and durations of landscape stability. The well-known soil-forming factors are presented as a theoretical framework for understanding landscape-scale soil evolution through time. Vertical and lateral patterns of stacked paleosols that appear in the rock record are discussed in order to address practical approaches to identifying and describing paleosols in the field. This chapter emphasizes a robust multi-proxy approach to paleopedology that combines soil stratigraphy, morphology, mineralogy, biology, and chemistry to provide an in-depth understanding of paleoecology.

Theoretical Background

Introduction

Soils are the product of organic matter accumulation and mineral weathering on the terrestrial surface, integrating the dynamic physical, biological, and chemical interactions that take place over time. Soils should not be confused with sediment, a material product of weathering and erosion occurring elsewhere and subject to transport by wind, water, or gravity. In contrast, soils form in situ and are the alteration products of weathered sediment, rock, and/or organic components. Dramatic changes in mineralogy and porosity accompany pedogenesis , which , along with bioturbation , act to disrupt primary depositional bedding. These processes impart new structural and chemical characteristics that are recognizable using field and lab techniques. Such patterns are preserved as soil profiles and can be buried and incorporated into the stratigraphic record (Fig. 9.1A). The resulting paleosols appear in many terrestrial depositional successions as intervals with variegated color, surface texture, and recessed erosional relief in outcrop.

Fig. 9.1

A, Schematic cross-section of a paleo-Critical Zone reproduced from Nordt and Driese (2013). Top Panel: Alluvial deposits in a continental-scale basin. Middle Panel: Regional basin with paleo-Critical Zones illustrated by colored lines that could be studied in outcrop or using sediment cores indicated by vertical black lines. Lower Panel: Local outcrop of an alluvial succession with a paleo-Critical Zones and paleosol. DTCI = Deep time critical interval. The black box indicates the location of Fig. 9.1B. B, A typical soil profile and typical paleosol profile preserved after burial and diagenesis . Modified from Tabor and Myers (2015)

State-Factor Approach

Paleopedology uses concepts and tools of Earth systems science to understand paleosols in the Quaternary and pre-Quaternary sedimentary record (Yaalon 1971; Retallack 2001; Sheldon and Tabor 2009; Tabor and Myers 2015). The conceptual framework of most paleosol-based paleoenvironmental studies relies upon the fundamental state-factor theory for modern soils (Jenny 1941):

$$ S\,or\,s = f\left( {cl,o,r,p,t, \ldots } \right), $$

(1)

where S is the soil and s is any soil property, and cl is climate, o is organisms, r is relief, p is parent material , and t is time. The series of dots, or ellipsis, represent other possible unaccounted for factors influencing S. For modern soils, Eq. 1 shows that the five primary factors (cl, o, r, p, t) influence the direction and magnitude of pedogenesis , S, or in some cases seem to predict S (Jenny 1941). For example, modern climosequence relationships that relate S to cl are applied to paleosols using an environmental-factor approach (sensu Retallack 1994), where s is measured in the paleosol using a variety of bulk and isotope geochemical, magnetic mineral, and physical proxies (Maynard 1992; Maher and Thompson 1995; Maher 1998; Ludvigson et al. 1998; Birkeland 1999; Retallack 2001; Sheldon et al. 2002; Holliday 2004; Driese et al. 2005; Dworkin et al. 2005; Retallack 2005; Nordt et al. 2006; Sheldon 2006; Sheldon and Tabor 2009; Nordt and Driese 2010a, b; Passey et al. 2010; Retallack and Huang 2010; Gulbranson et al. 2011; Gallagher and Sheldon 2013; Ludvigson et al. 2013; Hyland et al. 2015). The results of such soil-sequence studies allow for the development of empirical, quantitative relationships between soil properties and the state-factor studied (e.g., climate). The reader is directed to Rasmussen et al. (2005) and Rasmussen and Tabor (2007) and references therein for an example of alternative, quantitative frameworks for pedogenesis .

A Brief History of Paleosols in Paleoenvironmental Studies

The application of paleopedology as a tool for paleoecological reconstruction originated in the mid- to late 19^th century, when geologists used buried forest beds to correlate glacial stratigraphy in the Midwestern USA, and the early 20^th century, when many studies in the USA were focused on identifying the environments of “early man”, (i.e., Paleoindian culture) (Leighton 1937; Holliday 2004, 2006). The accumulated knowledge of the early to middle 20^th century likely led to the formal designation of the field as paleopedology at International Union for Quaternary Science (INQUA) meetings, led by pioneering paleopedologists Ruhe and Yaalon, which eventually led to two highly influential volumes (Ruhe 1965; Yaalon 1971).

The study of paleosols in the past 50 years has seen advancements through both pre-Quaternary and Quaternary disciplines. Some of the first pre-Quaternary investigations into paleoecology using paleosols can be traced to Retallack’s (1983) systematic study of Eocene-Oligocene paleosols in the Badlands of South Dakota, USA. Retallack’s (1983) work and successive studies ushered in an era of paleopedological research in a wide range of geologic settings (e.g., Bown and Kraus 1987; Bestland et al. 1997; Terry 2001). The science of paleopedology gradually evolved to become more quantitative with an increasing number of numerical proxies for climate, vegetation, and edaphic soil properties (Retallack 1994; An and Porter 1997; Maher 1998; Sheldon et al. 2002; Driese and Ober 2005; Dworkin et al. 2005; Retallack 2005; Prochnow et al. 2006; Kraus and Hasiotis 2006; Nordt and Driese 2010a, b; Retallack and Huang 2010; Gulbranson et al. 2011; Gallagher and Sheldon 2013; Trendell et al. 2013a, b; Hyland et al. 2015).

Paleosols are important paleoecological archives because they are records of Critical Zones of the Earth’s past, preserving the in situ substrate of ancient landscapes (Amundson et al. 2007; Leopold et al. 2011; Nordt et al. 2012, 2013; Marin-Spiotta et al. 2014). This approach to studying paleo-Critical Zones (paleo-CZs) continues to gain traction as a multi-disciplinary method for understanding the Earth’s environments and processes. Recent work has focused on unraveling and quantifying the biogeochemical processes operating within a paleosol system (Nordt and Driese 2010b; Nordt et al. 2012, 2013), comparable with modern Critical Zone studies (Fig. 9.1A; Amundson et al. 2007; Brantley et al. 2007). In light of these advances, this chapter establishes a theoretical framework for the field of paleopedology through the lens of a paleo-CZ and is applicable throughout geologic time.

This chapter does not review isotope geochemical methods in paleopedology and paleoenvironmental reconstruction, which can be found elsewhere (Nordt 2001; Sheldon and Tabor 2009; Cerling 2014; Levin 2015; Tabor and Myers 2015; Zamanian et al. 2016; Berke 2017). Much of the information compiled in this chapter was adapted from journal articles and books published by pioneers in the field of Quaternary and pre-Quaternary paleopedology (e.g., Yaalon 1971; Brimhall and Dietrich 1987; Birkeland 1999; Retallack 2001; Anderson et al. 2002; Holliday 2004; Sheldon and Tabor 2009; Nordt et al.2013; Schaetzl and Thompson 2015; Tabor and Myers 2015). The reader is encouraged to consult these publications, which contain conceptual and methodological frameworks for analyzing soils and paleosols.

Approaches

Macroscale Physical Characterization

Field observation is the first step for identifying and describing paleosols. This step relies upon observing paleosol morphological features with unassisted vision or a hand lens with ~10× magnification. Much like modern soils, paleosols are identified in the field using color, texture, and weathering features within peds or along former voids (i.e., ped/void features). A number of features may suggest the presence of a paleosol: authigenic mineralization, rhizoliths, coloration, bioturbation , presence of soil aggregates (peds), redistribution of primary/authigenic minerals, shrink-swell features, disruption of primary sedimentary bedding, and deflation surfaces or biomantles as indicated by stone lines. After the paleosol(s) has been identified, a trench should be excavated to a depth that exposes sediment or rock unaffected by modern processes.

Field description and horizon delineation: The description of some paleosols differs from soils due to the inhibiting effects of compaction and lithification on the measurement of many properties. We advocate for describing paleosols in as much detail as the geologic setting allows, as these data will aid in more informed interpretations of paleolandscapes, paleoclimate , and paleovegetation. The reader is directed to the Illustrated Guide to Soil Taxonomy (Soil Survey Staff 2014a) and Keys to Soil Taxonomy (Soil Survey Staff 2006) for standard protocols for field description and horizon designation.

Paleosol horizons should be described using a standardized, published descriptive nomenclature, such as U.S. Department of Agriculture (USDA) soil nomenclature (Tables 9.1 and 9.2; Schoeneberger et al. 2012). A typical soil profile includes the O, A, B, C and R master horizons , where O is the zone of concentrated organics, A is the zone of leaching, B is the zone of accumulation, C is unaltered or unconsolidated parent material , and R is consolidated parent material. The A and O horizons are typically not preserved or recognized in paleosols due to erosion or oxidation of organic matter during the burial process (Fig. 9.1B). Less common master horizons are described in Table 9.1. Stratigraphic successions preserved in the rock record consist of aggrading sedimentary deposits, and therefore many paleosols do not have R horizons.

Table 9.1 Master soil horizons that may be observed in paleosols and their paleoecological implications. Modified from Guthrie and Witty (1982), Soil Classification Working Group (1998), and Schaetzl and Thompson (2015)

Table 9.2 Suffix symbols used with master soil horizons and their possible paleoecological implications when observed in paleosols. Modified from Guthrie and Witty (1982), Soil Classification Working Group (1998), Soil Survey Staff (2010), and Schaetzl and Thompson (2015)

Subordinate indicators are added as lowercase suffixes to master horizons to identify the presence of pedogenic minerals, organic complexes, or other distinct features (Table 9.2). For example, common types of B horizons include Bk (accumulation of carbonate ) and Bt (accumulation of translocated clay; Fig. 9.2B, H). The naming of soil horizons begins with field description but is frequently modified after laboratory and micromorphologic analysis (Birkeland 1999).

Fig. 9.2

Photographs of pedogenic features in the field and in micromorphology. A, Oxidized and volcaniclastic-rich paleo-Aridisol from Olduvai Gorge , Tanzania. B, Close-up of lowermost step in trench in Fig. 9.2A, which has excellent horizonation, clay illuviation , and ped structure (angular blocky and columnar). C, Fining upward, coarse-grained alluvial paleosol from the Chalbi Desert, Kenya. D, Stacked paleo-Vertisols from Olduvai Gorge separated by a tuff with examples of blocky, platy, and wedge peds. E, FeMn filled rhizoliths from Olduvai Gorge (Reproduced from Beverly et al. 2014). F, Zeolitized rhizoliths from Olduvai Gorge (Reproduced from Beverly et al. 2014). G, Pedogenic B-fabric (Reproduced from Beverly et al. 2015a). H, Illuviated clay in a pore (Reproduced from Beverly et al. 2015a). I, Carbonate rhizolith and rhizocretions in thin section from Karungu, Kenya. J, Pedogenic carbonate in thin section with circumgranular cracking (Reproduced from Beverly et al. 2015a). K, Preserved root structures identified in a paleosol using UVf (Reproduce from Beverly et al. 2015b). L, Fecal pellets and meniscate backfill (Reproduced from Beverly et al. 2015a). M, Pedogenic slickensides in a paleo-Vertisol from Karungu, Kenya, and pedogenic carbonate nodule. N, Non-pedogenic carbonate from Olduvai Gorge . Groundwater carbonate precipitates after primary deposition. Note rock hammer for scale

Detailed field descriptions of paleosol profiles include texture, peds, mineralizations, color, the presence of fossil roots, trace fossils and any other notable features (e.g., Schoeneberger et al. 2012). Paleosol texture is equivalent to grain size in sedimentology and is described in the field or measured in the laboratory. Peds should be identified and described with respect to size, shape, and grade (Fig. 9.3; Table 9.3; Schoeneberger et al. 2012). Ped types are related to parent materials , intra-profile water cycling, biotic activity, and degree of pedogenic alteration, all of which have net paleoecological implications (Table 9.3). For example, columnar peds, which are vertically oriented with round tops, form as a result of clay flocculation in saline soils of arid and semi-arid environments (Fig. 9.3). Such soils tend to exclude most plant types due to the combined effect of low porosity, which inhibits root growth, and nutrient limitation (Qadir and Shubert 2002).

Fig. 9.3

Photographs of common peds (soil aggregates) identified in paleosols. See Table 9.3 for explanation of these ped types

Table 9.3 Classification of soil peds or aggregates. Modified from Retallack (2001) and Brady and Weil (2008). Examples are shown in Fig. 9.3

Recognition of the pedogenic redistribution of primary minerals and precipitation of secondary minerals can offer support for interpretations of climate, redox conditions, or soil maturity. Clays are translocated downward with soil water from eluvial (leaching) to illuvial (accumulating) horizons during wetting events. Subsequent drying causes the suspended clays to adhere to void features, which include the surfaces of peds, slickensides, root traces, and tubules. Calcium carbonate precipitates along root traces as rhizoliths or as discrete nodules in the matrix of Bk horizons. Similarly, Fe–Mn masses precipitate along voids or as hard matrix nodules as a result of alternating redox conditions. Salts, such as barite or gypsum, typically accumulate as discrete masses in the matrix of soils with extreme water stress. Geochemical and mineralogical approaches to the characterization of pedogenic minerals are addressed below.

The color of the paleosol matrix, mineralizations, and other notable features are characterized using Munsell^® soil color charts. Matrix color can be used to identify soil organic matter (SOM) input and iron speciation associated with redox states but should only be used as an initial tool for description before further laboratory analysis (Holliday 2004). Darker colors, such as the dark browns or dark grayish browns found in buried A horizons, may suggest higher SOM production. However, dark colors can also reflect parent materials , especially for Mg-smectite clays (Fig. 9.2D; Beverly et al. 2014, 2015) or mafic igneous rocks.

The oxidation state and mineralogy of Fe imparts distinct color characteristics on soils and paleosols (Schwertmann 1993; Birkeland 1999) and may directly reflect local environmental conditions such as drainage, fire, temperature, and duration of soil development (Table 9.4; Holliday 2004). Green, blue, or gray colors often suggest poorly-drained, anaerobic conditions (Vepraskas 1992, 2001). Prolonged saturation reduces the amount of available oxygen for soil respiration, which leads to microbially-mediated reduction of Fe and Mn species and an overall gray appearance (Vepraskas and Faulkner 2001). Re-precipitation of Fe and Mn in localized oxygen-rich zones creates a mottled appearance (Fig. 9.2C). These processes frequently occur in wetland or lake-margin paleosols proximal to paleo-lakes or on paleo-floodplains (Retallack 1994; Ufnar et al. 2004; Tabor et al. 2007; Rosenau et al. 2013; Ashley et al. 2013, 2014; Beverly et al. 2014; Driese and Ashley 2016).

Table 9.4 Paleoecological implications of soil color. Modified from Schwertmann (1993) and Holliday (2004)

Red or reddish-brown horizons usually indicate well-drained conditions or higher temperature soil-forming environments (Fig. 9.2A–C; Ashley and Driese 2000; Driese et al. 2016). Well-drained conditions promote Fe oxidation in O₂-rich soils, which leads to the development of goethite and hematite that impart a reddish-brown to red color in the soil (Fig. 9.2A–C; Schwertmann 1993; Birkeland 1999; Holliday 2004). Increasing temperatures can facilitate chemical reactions, and this is thought to influence the reddening of the soil. Fire-altered soil often shows reddening (Schwertmann 1993; Mentzer 2014) like those affected by heating from lavas. However, careful observation of pedogenic features and geochemical trends can disentangle the competing roles of pedogenesis versus thermal alteration in causing soil redness. For example, Sheldon (2003) showed evidence that rubified paleosols between Columbia River basalt flows appear red primarily due to oxidation and clay production during pedogenesis rather than thermal alteration. Paleosols can also inherit their color from parent materials , so care must be taken when interpreting the paleoclimatic or paleoenvironmental significance of red paleosols (Gile 1979; Sheldon 2005). Finally, Maxbauer et al. (2016b) and Clyde et al. (2013) recently showed that oxidation has altered paleosol color to depths of 20 to 30 m below the modern outcrop surface in core from the Bighorn Basin Coring Project. In the unweathered portion of the core, the colors are much more muted in comparison to the weathered paleosols in outcrop and in the upper 25 m of the core.

Sampling of paleosol horizons: After paleosol horizons are identified and their thicknesses and properties have been described, bulk soil , sediment, or paleosol samples should be collected by horizon. It is important to collect “unweathered” rock sample and avoid modern fractures and weathering (see Maxbauer et al. 2016b). Soil micromorphology (see below) can be used to identify pedogenic processes, evidence of overprinting, and mineralogical composition. Undisturbed, oriented thin-section samples should be collected from all representative horizons, air-dried, epoxy impregnated, and then prepared for in-house or commercial fabrication. In unlithified paleosols, epoxy impregnation may be necessary in the field for removal of oriented samples because of their unconsolidated state.

Bulk samples can also be collected using an arbitrary sampling interval (e.g., every 10 cm), depending on the research question(s) that are being addressed. The bulk samples can be analyzed for their particle size (depending on degree of lithification), bulk and isotopic geochemistry, and bulk and clay mineralogy. Samples of unique features should also be collected if time and space permit, e.g., root traces, mineralizations, diagnostic ped/void features (see below), organic mats, and trace fossils . Soil trace fossils can be used as a complementary data set to paleosol analysis. For further discussion of ichnology see Hasiotis (2007), Hasiotis et al. (2007a, b), and Hembree et al. (2014).

Bulk samples for geochemical analysis should be collected from uppermost B horizons (ideally 25–100 cm depth), as required for some paleoclimate reconstructions (Sheldon et al. 2002; Nordt and Driese 2010a, b; Stinchcomb et al. 2016). Bulk density of paleosols is required for constitutive characterization (e.g., mass-balance geochemistry; see section on Constitutive Mass-Balance). The bulk density of soil and sediment samples can be calculated using the clod method on peds or core method if unconsolidated (Blake and Hartge 1986). If this is not possible, bulk density can be estimated using pedotransfer functions based on the organic content and texture (Rawls 1983) or SiO₂ concentration in Vertisols (Nordt and Driese 2010b).

Micromorphological Characterization

Soil micromorphological features can be studied using a polarized light microscope and described using a variety of terms developed for micromorphological observation of soils. Brewer (1976) is the seminal work regarding the theoretical framework and application of soil micromorphology for interpreting soil development. Later, simpler terminology was developed by Bullock et al. (1985), Fitzpatrick (1993), and Stoops (2003). Micromorphological terminology can be extensively subdivided and modified depending on the data needed to answer specific research questions. An analysis of the microstructure often begins with a description of the mineral soil matrix by describing the coarse to fine ratio (c/f) and oriented birefringent fabric (b-fabric). A description of weathering of primary minerals can be useful for determining the degree of pedogenic development (Bullock et al. 1985). B-fabric within the soil matrix is often described and can be indicative of the paleo-hydrology of the system. For further discussion of b-fabrics see Stoops (2003) and Stoops et al. (2010). Pedogenic slickensides can be identified from illuviated clay by a lack of laminations and an identical grain size between the matrix and oriented clay (Fig. 9.2G). Oriented b-fabric may indicate that repetitive wetting and drying caused clays to shrink and swell, leading to inclined orientation due to the confining pressure of overlying soil (Blokhuis et al. 1990). It could also indicate bioturbation and reorganization of the soil matrix by other means creating grano-, channel, or poro-striated b-fabric (Stoops et al. 2010).

Pedofeatures, which are >20 μm and are distinguishable from the soil matrix, should also be described in detail (Stoops 2003). In thin section, burrows and root voids can be distinguished by analyzing the textures around the void (Stoops 2003; Stoops et al. 2010). Coatings (lining voids), hypocoatings (impregnated into matrix around pore), quasicoatings (impregnated into matrix surrounding pore but at some distance from feature), and infillings are also important to describe. A variety of materials such as Fe–Mn (Fig. 9.2E), illuviated clay or silt (Fig. 9.2H), or carbonate can fill these voids. Both silt and clay can be translocated down profile via the vertical drainage of water (Stoops et al. 2010) and indicate that the water table was low enough for drainage, but that there was enough water to translocate clay or silt. The illuviation of silt in thin section is sometimes used to infer the physical translocation of freshly deposited silt after a flood or dust accumulation (Kubiena 1970; Fedoroff and Goldberg 1982; Kemp 1999).

The accumulation of clay coatings observed in thin section has been related to the residence time (duration of weathering in the system) of soils using a chronosequence of deeply weathered, sub-tropical soils from Mississippi (Ufnar 2007). It is possible that the Ufnar (2007) chronofunction may overestimate the length of pedogenesis in certain climates (i.e., monsoonal), but this technique can be a useful approximation (Beverly et al. 2015a). Although clay illuviation has been inferred to represent warm-wet climate conditions in some studies (e.g., Bronger and Heinkele 1989), later work showed that Bt horizons can form in semi-arid grassland landscapes when enough time has elapsed (Han et al. 1998). A corollary of the Han et al. (1998) findings is that clay coatings could result from windblown dust deposits weathered and translocated down from silty A horizons.

Fe–Mn concentrations and depletions can be used to infer drainage (Fitzpatrick 1984; Bullock et al. 1985) and more specifically to infer the relative influence of epi- versus endo-saturation in the paleosol. Fe oxides coating peds with reduced interiors are indicative of endosaturation, where water is wicked upward through micropores within aggregate interiors. Oxidizing conditions remain in macropores, where the spacing between grains is too great for water to adhere. Reduction present along the exterior of peds that contain oxidized interiors suggests periodic episaturation (Fitzpatrick 1984; Bullock et al. 1985). Water saturation from surficial ponding will fill macropores and inhibit the diffusion of atmospheric oxygen into pore waters. Consumption of oxygen by microbes will result in the reduction of Fe on aggregate exteriors, which will penetrate into aggregate interiors with time. If the soil becomes drained of ponded water before the anoxic pore waters completely infiltrate aggregate interiors, Fe within aggregates will remain oxidized (Fitzpatrick 1984; Bullock et al. 1985).

Nodules are three-dimensional features not associated with a void or pore feature and are important as they can reveal information about calcification or redox conditions (Stoops 2003). The most commonly identified nodules are pedogenic carbonate , but Fe–Mn nodules are also important (Stiles et al. 2001). The accumulation of CaCO₃ in soils frequently indicates an arid or semi-arid soil-forming environment (Gile et al. 1966; Machete 1985). Cathodoluminescence (CL) is also a useful tool for determining the redoximorphic environment of carbonate identified in a paleosol, assessing diagenetic alteration (Barnaby and Rimstidt 1989; Machel et al. 1991), or identifying parent material contamination by marine carbonates (Michel et al. 2013). Carbonate forming in a shallow reducing environment with higher manganese to iron ratios will luminesce; whereas, carbonate forming in an oxidizing environment will have no luminescence (Barnaby and Rimstidt 1989; Machel et al. 1991). The importance of pedogenic carbonate is further discussed below.

An ultraviolet fluorescence (UVf) microscope attachment causes organic matter to fluoresce, facilitating the identification of organics that would be otherwise unobservable. UVf can reveal features such as root cell structure not visible in cross-polarized or plane light (Fig. 9.2K; Beverly et al. 2014, 2015b). Thin sections are often point counted to quantify mineralogy, organic content, or other morphological features that elucidate pedogenic processes (e.g., Stinchcomb et al. 2014).

Identifying features created by soil fauna often yields critical paleoecological information. Faunal features include: fecal pellets, channels or burrows , chambers, coatings, and infillings (Stoops 2003). Faunal features are not always readily identifiable but can yield significant information about paleohydrology. Termites, for example, do not live below the water table, and thus the depth of their burrowing can indicate the height of the water table associated with their colonization surface. Soil-dwelling crayfish can be of similar use in identifying the water table because they require standing water at the base of their burrows (Hasiotis and Honey 2000). Earthworm fecal pellets and burrows are often identified in paleosols and provide macropores favorable to root growth and microbial activity and stabilize organic carbon in the soil (Fig. 9.2L; Stoops et al. 2010). For further discussion of faunal features in micromorphology see Ch. 18 of Stoops et al. (2010).

Characterizing Pedogenic Minerals

Bulk mineralogy of soils and paleosols can be characterized using X-ray diffraction (XRD) (Poppe et al. 2001). Common pedogenic minerals and their paleoecological implications are shown in Table 9.5 (Bullock et al. 1985). Harris and White (2008) provide a summary of how to use XRD to identify common minerals in soils.

Table 9.5 Common pedogenic minerals and their paleocological implications. Modified from Bullock et al. (1985)

Clay Mineralogical Characterization: Clay mineralogical analysis is a powerful tool for recognizing weathering products versus inherited materials, which ultimately aids in understanding the weathering pathway(s) that occurred in paleosols (e.g., Tabor and Montañez 2002; Rosenau et al. 2013). The composition of the clay-sized fraction (<2 µm) is analyzed using XRD and should be prepared following standard methods (Moore and Reynolds 1997). When preparing samples for clay mineralogy, multiple aliquots may be needed to help diagnose 2:1 clay minerals using the following treatments: (1) air-dry, (2) K saturation, (3) Mg saturation by ethylene glycol solvation , and (4) subsequent 550°C heat treatment of the K-saturated sample. Tabor and Myers (2015) address the characterization and interpretation of clay minerals for paleoclimatic purposes.

Pedogenic Carbonate: Pedogenic calcium carbonates (CaCO₃ – limestone ; (Ca,Mg)(CO₃)₂) – dolomite) are an important paleoecological indicator as they typically form in dry environments where precipitation is exceeded by evapotranspiration (Fig. 9.2K), though carbonate precipitation can occur in some poorly-drained alluvial back-swamp soils (Aslan and Autin 1998; Mintz et al. 2011). Carbonates initially precipitate in the soil matrix as soft masses with diffuse boundaries and micritic crystalline texture (Nordt et al. 2004). Zones of carbonate can develop septarian shrinkage cracks (Fig. 9.2J), multi-generational carbonate overgrowths, and coatings of other minerals, such as Fe and Mn oxides (Bullock et al. 1985; Stoops 2003). Such alteration converts the soft masses to hard nodules, which has been interpreted to occur outside the zone of active carbonate precipitation (Nordt et al. 2004). Carbonates also precipitate locally around roots, where CO₂ is high due to respiration, forming filaments and nodules that converge to form rhizoliths. If vertic soil processes caused by shrinking and swelling of clays during the wet and dry seasons are present, these rhizoliths can be broken up to form rhizocretions (Fig. 9.2I).

The degree to which carbonates precipitate in the soil matrix is primarily a function of time (provided that the other soil forming factors are held constant) and can be described in stages of accumulation (Table 9.6; Gile et al. 1966; Machete 1985). Soil carbonates can also be inherited from the parent material , particularly when materials are sourced from marine limestone (Michel et al. 2013). When describing carbonates in the field, an effervescence test is useful for describing the degree of effervescence in the paleosol matrix and/or in secondary carbonate features, which may indicate the relative concentration of carbonate in the paleosol horizon (Schoeneberger et al. 2012).

Table 9.6 Stages of calcium formation in calcic soils and pedogenic calcretes. Modified from Gile et al. (1966) and Machete (1985)

Stable isotopes of carbon and oxygen from pedogenic carbonates have been used extensively to interpret paleoecology (Cerling and Hay 1986; Cerling et al. 1988), where long sedimentary records have been used to understand the changing climate and environment of the past several million years. Carbon isotopes in pedogenic carbonates are of particular importance because they can be used to interpret open and closed environments by using the enrichment or depletion of ¹³C relative to ¹²C, which can result from changes in biomass utilizing disparate photosynthetic pathways (C₃ vs C₄) or from water stress in C₃ plants (Cerling and Quade 1993). Pedogenic carbonates should be collected from at least 50 cm depth to ensure that the soil CO₂ reflects the ¹³C composition of respired organic C in soil CO₂ rather than infiltrated atmospheric CO₂ (Cerling and Quade 1993). A discussion of biases associated with pedogenic carbonate can be found in Biases and Shortcomings.

Pedogenic Siderite: Siderite (Fe(II)CO₃) crystallizes in the reducing phreatic waters of poorly drained soils, commonly around the roots of hydrophilic vegetation (Ludvigson et al. 1998, 2013). Spherulitic masses of siderite are detectable in thin-section as typically mm-scale crystals with radial extinction. In addition to its use as a drainage and redox indicator, siderite has been used to reconstruct the δ¹⁸O composition of ancient meteoric water (Ludvigson et al. 2013, and references therein).

Geochemical Characterization

In contrast to mineralogical characterization, which reveals the organization of elements in soils, geochemical analysis measures the concentrations of elements in a sample. Paleosols should be geochemically analyzed to determine parent material uniformity, the presence and magnitude of pedogenic processes (Birkeland 1999), and to estimate paleoenvironments or paleoclimate variables (Sheldon and Tabor 2009; Tabor and Myers 2015). Bulk geochemical samples should be prepared in a similar fashion to modern soils. Samples are air-dried, broken up with a mortar and pestle, and macroscopic minerals, such as mineral nodules or salts, are discarded so that only the soil matrix is analyzed (Soil Survey Staff 2014b). Treatment with HCl is not necessary and will bias quantitative geochemical analysis if Ca²⁺ in the matrix is removed with carbonate (Myers et al. 2014; Tabor and Myers 2015). Samples are then powdered (typically with >85% of sample finer than 100 μm) and analyzed using inductively-coupled plasma (ICP) – mass spectrometry (MS), – optical emission spectrometry (OES), – atomic emission spectrometry (AES), or X-ray fluorescence (XRF). The results are usually reported in elemental or oxide weight percent. Several applications of geochemical characterization are described here, but for a more thorough review, see Sheldon and Tabor (2009) and Tabor and Myers (2015).

Molecular Weathering Ratios: Molecular weathering ratios can be used as a qualitative approximation to understand weathering in the paleosol and to evaluate relative changes down profile (Retallack 2001), but more quantitative methods such as constitutive mass-balance are now preferred by paleopedologists (see below). Oxides are by convention converted from weight percent to molar percent (dividing by molecular weight of each oxide) before calculating molecular weathering ratios. A discussion of the inferences that can be drawn from molecular weathering ratios can be found in Retallack (2001). The chemical index of alteration (CIA) and chemical index of alteration minus potassium (CIA-K) can be used as a proxy for weathering intensity in a paleosol (Table 9.7; Nesbitt and Young 1982; Maynard 1992; Sheldon et al. 2002). These weathering indices measure clay formation and base loss associated with feldspar weathering and base retention and carbonate precipitation in arid environments (Maynard 1992; Sheldon et al. 2002; Lukens et al. 2018). CIA-K is the preferred weathering ratio for pre-Quaternary paleosols as it reduces digenetic effects of potassium metasomatism (Maynard 1992; Sheldon et al. 2002).

Table 9.7 Common MAP proxies

Constitutive Mass-Balance: Mass-balance is a more powerful tool than molecular weathering ratios because it quantitatively compares geochemical and volumetric changes in soil horizons relative to their parent material (Brimhall and Dietrich 1987; Brimhall et al. 1991a, b). Due to variability in parent material compositions, the first step in the mass-balance approach is to evaluate parent material composition using immobile and chemically inert major, trace, and rare earth elements (e.g., Ti, Zr, Al, Nb, La, Ce). The most commonly used immobile constituents are Ti, Zr, and Al, as the deviation from the parent material tends to be relatively small. The concentration of other elements (e.g., Na, Mg, K, Ca, Si, Fe(II), Mn(II), C, P) is weighed against the immobile element to elucidate additions and/or losses relative to the parent material. For these methods, selection of an appropriate, homogeneous, and unweathered parent material is best, although some studies have found that using the lowermost, least-weathered horizon for each paleosol provides acceptable results (e.g., Driese et al. 2016). If the parent material is approximated, it is best to discuss the mass-balance results in a conservative manner (e.g., minimum apparent loss or gain of an element).

Vertical changes in grain size in soil profiles are common and induce heterogeneity in starting mineralogy, which can influence the results of mass balance calculations. Likewise, additions of sedimentary material during pedogenesis can produce variability in weathering trends toward the top of profiles. To account for this issue, parent material uniformity can be examined using geochemical reconstructions applied to buried soils and paleosols following Maynard (1992). The ratio of TiO₂/Zr for each candidate parent material and the ratio TiO₂/Zr of the average composition of the paleosol can be used to determine percent deviation from the parent material (Maynard 1992):

$$ {\text{Percent}}\;{\text{Deviation}} = {\left( \frac{{\text{Ti}}_{\text{soil}}}{{\text{Zr}}_{\text{soil}}} - \frac{{\text{Ti}}_{\text{parent}}}{{\text{Zr}}_{\text{parent}}} \right)\big/\left( \frac{{\text{Ti}}_{\text{parent}}}{ {\text{Zr}}_{\text{parent}}} \right) \times 100} $$

(2)

Cross plots of immobile elements are also useful for determining parent material such as Zr/(Zr/TiO₂) vs. TiO₂/(TiO₂/Zr) (e.g. Ashley and Driese 2000; Driese et al. 2000; Beverly et al. 2014). The selection of Ti and Zr as immobile index elements has also been discussed by Stiles et al. (2003a, b), who found that Zr can be introduced as windblown silt in semi-arid environments.

The following calculations for constitutive mass-balance analysis are derived from Brimhall and Dietrich (1987), Brimhall et al. (1991a, 1991b), Chadwick et al. (1991) and Anderson et al. (2002). Closed system weathering conserves mass and volume, such that:

$$ {\text{V}}_{\text{p}}\uprho_{\text{p}} {\text{C}}_{{{\text{i}},{\text{p}}}} = {\text{V}}_{\text{w}}\uprho_{\text{w}} {\text{C}}_{{{\text{i}},{\text{w}}}} $$

(3)

where the volume (V), bulk density (ρ) and concentration (C) of an immobile element (i) in the parent material (p) are equal to those of a weathered horizon (w).

In most soils, weathering is an open system process, wherein constituents are added or removed from the system. In such cases, volumetric change is tracked as strain (ε):

$$ \varepsilon_{i,w} = \frac{\rho_{p} C_{i,p}} {\rho_{w} C_{i,w}} - 1 $$

(4)

Note that if losses of an element in a weathered horizon are offset by changes in volume, strain is zero. Elemental additions and losses are calculated using the open-system mass-transport function (τ), which incorporates volumetric changes:

$$ \uptau_{j,w} = \frac{\rho_{w} {C}_{j,w}} {\rho_{p} {C}_{j,p}} \left( {\upvarepsilon_{i,w} + 1} \right) - 1 $$

(5)

where C_j,w is the concentration of an element of interest (j) in weathered material and C_j,p is the concentration of an element of interest (j) in the parent material . When τ_j,w = −1, 100% of the element j has been removed in a weathered horizon compared the parent material. Likewise, when τ_j,w = 1, 100% of element j has been added to a weathered horizon compared the parent material starting composition. Definitions of all variables in the above equations are summarized in Table 9.8.

Table 9.8 Glossary of variables modified from Brimhall and Dietrich (1987)

Interpretation of mass-balance calculations should be formed in terms of soil-forming process, which may vary depending on the environmental setting or parent material type. Translocation of carbonate is identified by simultaneous changes in Ca, Mg, and Sr. Clay or zeolite accumulation is indicated by increases in Na, Al, Si, and Mg (Ashley and Driese 2000; Beverly et al. 2014). In contrast, Mg fluxes have also been found to closely track clay-rich, Mg-smectite parent material rather than carbonate (Beverly et al. 2014). Losses of the redox-sensitive elements (e.g., V, Fe, Cu, and Mn) may indicate prolonged waterlogging followed by drainage. Cu is also important in the formation organic ligands and can be indicative of changes in organic matter content (Brantley et al. 2007). Likewise, phosphorus enrichment at the soil surface can indicate biocycling (Brantley et al. 2007). Human remains and debris can also be enriched in P relative to surrounding material, but additional evidence would be needed to support this interpretation (Eidt 1985; Holliday 2004; Holliday and Gartner 2007).

Pedotransfer Functions: Pedotransfer functions relate bulk geochemistry of soils to properties that are not directly measureable in paleosols. For example, modern soils are studied using a suite of standard characterization properties, which are measured on unconsolidated material to assess physical and chemical properties operating at the colloidal scale (Soil Survey Staff 2014b). Lithification and compaction of paleosols precludes such measurements, but the relationship between characterization properties and geochemical indices allows for detailed pedologic study of paleosols (Nordt and Driese 2010b). Soil characterization properties can yield quantitative paleoecological information on paleosol fertility and hydrology unavailable by any other means (Table 9.9; Nordt et al. 2011). A detailed explanation of modern soil properties can be found in Soil Survey Staff (2014b). Currently, most pedotransfer functions are restricted to paleo-Vertisols (Nordt and Driese 2010b), which are commonly found in the geologic record (Driese and Forman 1992; Driese et al. 1993; Prochnow et al. 2006; Cleveland et al. 2007; Rosenau et al. 2013; Torres and Gaines 2013; Driese et al. 2016). Efforts to develop more broadly applicable pedotransfer functions for soil characterization variables are ongoing and currently limited to soil pH (Lukens et al. 2018). If a paleosol is unconsolidated, soil properties could theoretically be directly measured using established methods (Nordt et al. 2013; Soil Survey Staff 2014b). However, it is not clear to what extent characterization properties are affected in buried paleosols that have not experienced macroscopic evidence of diagenesis .

Table 9.9 Pedotransfer functions for paleo-Vertisols modified from Nordt and Driese (2010b)

Mean Annual Precipitation and Temperature Proxies

Mean annual precipitation (MAP) can be estimated from paleosols by either physical or geochemical techniques (Table 9.7). The depth to calcic horizon (Bk) and depth to gypsic horizon (By) in soil profiles increases as a function of increasing MAP (Retallack 2005; Nordt et al. 2006; Retallack and Huang 2010). Nordt et al. (2006) used a “family of curves” approach to derive significant relationships between depth to nodular Bk horizons based on the concentration of nodules (1, 2, and 5%) for each of the microhigh and microlow portion of Vertisol profiles. The limitations for applying these depth-to-mineral functions are: (1) the tops of paleosols are commonly eroded (Myers et al. 2014), which results in MAP underestimates, (2) carbonate and gypsum must be nodular in morphology, (3) carbonate and gypsum must be analyzed in thin section to demonstrate pedogenic origin, and (4) burial compaction must be accounted for (Sheldon and Retallack 2001).

The bulk geochemistry of paleosol B horizons is a more reliable method of reconstructing MAP because it does not require the original thickness of the paleosol to be preserved. Typically, the uppermost B horizon is used as the “control interval” (25–100 cm depth in profile; Nordt and Driese 2010a). Bulk geochemistry is related to MAP through two processes: (1) hydrolysis of silicate minerals and progressive leaching of base cations at high MAP, and (2) precipitation of carbonates and retention of base cations at low MAP. Sheldon et al. (2002) found the weathering ratio CIA-K (Sect. 4.5.1) to be the most useful predictor of MAP across a range of soil types, which increases as CaO and Na₂O are preferentially leached relative to Al₂O₃. The acid neutralization capacity of secondary carbonates precipitated at low MAP (sensu Chadwick and Chorover 2001) results in retention of structural CaO and Na₂O and correspondingly low CIA-K values. K₂O is excluded from CIA-K because of potential diagenetic addition of K⁺ during illitization (Fedo et al. 1995; Sheldon et al. 2002; Lukens et al. 2018).

More recently, attempts have been made to decrease standard error in MAP functions by developing proxies specific to more narrow ranges of soil environments (i.e., cl, o, r, p, t; Sect. 3.1). CALMAG (Nordt and Driese 2010a) was developed specifically for clay-rich Vertisols, which are soils dominated by vertic features such as gilgai topography, pedogenic slickensides, and wedge peds that develop in soils that experience seasonal water deficit. Vertisols tend to be developed on pre-weathered parent material where little hydrolysis occurs, and therefore, tracking the flux of CaO and MgO (relative to Al₂O₃) as carbonates or exchangeable cations within the system results in improved MAP prediction (Nordt and Driese 2010a).

It should be noted that climofunctions using weathering indices that relate base cations to Al₂O₃ are only useful for estimating MAP up to a maximum of ~1600 mm (Sheldon et al. 2002; Nordt and Driese 2010a). Above 1600 mm, kaolinite is the dominant clay mineral and any weatherable bases are either completely depleted – yielding an effective ratio of alumina:alumina that does not change with increasing MAP – or bases have yet to be leached and the soil is not in equilibrium with climatic conditions. A few novel methods to estimate MAP utilize Fe in Fe–Mn nodules (Stiles et al. 2001) or the relative abundance of goethite to hematite (Hyland et al. 2015). The weight percent of Fe in Fe–Mn nodules found in Vertisols increases directly with MAP (Stiles et al. 2001) and can be used in conjunction with other methods. Most recently the goethite to hematite ratio (G:H) in the B horizon of a global dataset of soils is shown to correlate with MAP (Hyland et al. 2015). The ratio of G:H can be measured using either X-ray diffraction (XRD) or isothermal remnant magnetization (IRM) acquisition curves, but IRM is preferred due to the higher sensitivity in low concentrations of magnetic minerals in soils and paleosols (Hyland et al. 2015). G:H is the first MAP proxy to extend beyond the barrier of 1600 mm, with a predictive range of 100–3300 mm (Hyland et al. 2015). However, Maxbauer et al. (2016a) note that goethite is difficult to measure using IRM methods and easily recrystallizes during diagenesis , which would preclude measurements of original G:H ratios for most paleosols.

Mean annual temperature (MAT) proxies for paleosols have also improved over the last decade. Early attempts relied on bulk geochemical proxies such as the salinization index (Sheldon et al. 2002). An alternative approach to predicting paleo-MAT uses the oxygen stable isotopic composition of carbonates (δ¹⁸O ), which is dependent on both water (δ¹⁸O) composition and the temperature at which carbonates precipitated (i.e., soil temperature) (Cerling 1984). Dworkin et al. (2005) found a possible way to side-step original water composition by combining two equations that relate (1) the fractionation of oxygen from meteoric water into carbonate, and (2) the spatial relationship between MAT and the δ¹⁸O of meteoric water. However, it is possible that past time intervals had systematic differences in the spatial distribution of meteoric water δ¹⁸O and variation in mean annual temperature, which would preclude the application of this proxy (e.g., Lukens et al. 2017a).

Recent advances in the analytical resolution of isotopes of carbon and oxygen have allowed for the measurement of clumped isotopes (Δ₄₇). In this technique, the abundance of a doubly substituted CO₂ molecule of mass 47 (¹³C¹⁸O¹⁶O) produced when carbonate is digested during measurement is proportional to soil temperature, without assumptions based on original δ¹⁸O composition of soil waters (Ghosh et al. 2006; Passey et al. 2010; Quade et al. 2013). These carbonates are probably recording the warm, dry season temperature when carbonate is more likely to form and therefore do not represent a mean annual temperature (Passey et al. 2010; Gallagher and Sheldon 2016). The Δ₄₇ temperature estimates provide a unique view into critical paleoclimatic variables recovered from paleosols (e.g., Snell et al. 2013) when burial temperatures do not exceed 100°C (Henkes et al. 2014).

Classifying Paleosols

Several perspectives exist on the taxonomic classification of paleosols (Mack et al. 1993; Retallack et al. 1993; Nettleton et al. 2000), which stem from the fact that some distinctions in modern soil taxonomy rely on soil and climate properties (e.g., moisture and temperature regimes) that are immeasurable in paleosols. Each system has strengths and weaknesses (i.e., Retallack et al. 1993, comment and reply), and to date no taxonomy is uniformly accepted among paleopedologists. The two most commonly applied paleosol classification systems are Mack et al. (1993) and Retallack et al. (1993). The Mack et al. (1993) system classifies paleosols without the use of modern soil properties such as bulk density or base saturation, which are not measurable in most paleosols, and creates new names for some paleosol orders. The Retallack et al. (1993) classification system simplifies USDA Soil Taxonomy to allow for paleosol classification by omitting properties that cannot be directly measured in paleosols but retains the naming conventions of Soil Taxonomy. The authors prefer to adapt Soil Taxonomy where possible for comparisons with modern soils and to encourage collaboration with modern soil scientists. The Illustrated Guide to Soil Taxonomy (Soil Survey Staff 2014a, b) is a useful resource for those unfamiliar with the complexities of Soil Taxonomy and is practical for most paleopedological classification purposes.

Rates of Pedogenesis: Paleosols and paleosol successions can also be categorized based on the varying degrees of pedogenesis and sedimentation. Paleosols in alluvial environments are broadly split into three different categories : compound, composite, and cumulative (Fig. 9.4). Compound soils are weakly developed with vertically stacked profiles that occur when sedimentation rate is not steady and there is little erosion . In composite or welded paleosols the rate of pedogenesis exceeds that of sedimentation, and with little erosion creates vertically stacked profiles that partially overlap. Cumulative paleosols have slow but overall constant sedimentation rates where pedogenesis can modify the sediment as it is deposited (Kraus 1999). Cumulative soils tend to preserve remains, such as artifacts or fossils, in discrete depth ranges (Holliday 2004, and references therein). With erosion, these three categories can be further split into compound truncated set, composite truncated set, and cumulative truncated set (Kraus 1999). A careful study of root traces can also help classify rates of pedogenesis . Root mats can help identify surfaces within the paleosol but can also form at a lithologic boundary or fragipan (x horizon suffix; Table 9.3) preventing growth. Clay coatings, redoximorphic features, carbonate , Fe–Mn, and zeolites can all preserve roots features (Fig. 9.2E–F; Ashley and Driese 2000; Stoops et al. 2010; Beverly et al. 2014, 2015a). For further discussion of rhizoliths and their paleoecological implications see Kraus and Hasiotis (2006).

Fig. 9.4

Reproduction of Fig. 9.2 from Kraus (1999) illustrating the pedofacies method of categorizing paleosols. Abbreviations: Ag – gleyed A horizon; Bg – gleyed B horizon; Bw – B horizon showing color or structure development but little if any illuvial accumulation; Bt – B horizon showing accumulation of clays; Btj – incipient development of a Bt horizon; Cg – gleyed C horizon. Reprinted with permission from Elsevier

Pedofacies and Pedotypes: Paleosols in stratigraphic successions often contain similar attributes, which relate to their depositional environment. Such clustering can reveal vertical and lateral trends that are the result of geologic and climatic phenomena. Changes in paleosol maturity and drainage are typically assessed using qualitative indices (Table 9.10). Pedofacies use a combination of sedimentological and pedological features to categorize paleosols (Bown and Kraus 1987; Kraus 1987). The pedofacies approach was developed in alluvial settings and relies on the inverse relationship between deposition and soil development – grain size and flooding frequency decrease with distance from sediment source, which is accompanied by increased soil maturity and thicker individual soil profiles.

Table 9.10 Paleosol maturity and drainage indices. Modified from Atchley et al. (2013)

Retallack (1994) developed a more universal model of paleosol categorization, in which paleosols with similar features are grouped as pedotypes . A pedotype is a representative paleosol and reference to the type profile for each pedotype allows for a non-genetic classification of each paleosol and is indicative of one set of soil forming (cl, o, r, p, t) conditions (Retallack 1994). Analysis of pedofacies and pedotypes in vertical stratigraphic successions can reveal long-term trends in climate or basin subsidence (Atchley et al. 2004; Cleveland et al. 2007; Aziz et al. 2008; Atchley et al. 2013). By modifying concepts originally developed in marine sequences, Atchley et al. (2004, 2013) used combinations of grain size and paleosol thickness, maturity, and drainage to delineate fluvial aggradational cycles (FACs) where fining upward successions are overlain by paleosols (e.g., Figure 9.2C). See Atchley et al. (2013) for a review of paleosol stacking pattern analysis.

Strengths of Approaches

Because paleosols are by their nature an unconformable surface, they have been ignored in the past in favor of sediment cores from lakes or oceans to understand the effect of long-term climate changes on paleoecology . However, paleosols provide regional, site-based, environmental reconstructions and are thus very useful for understanding paleoecological variability in the immediate site area. Collecting paleosol samples for analysis using the variety of methods previously mentioned is fairly simple. When used properly to reconstruct climate or vegetative cover, paleosols act as in situ archives that record environmental conditions on moderate time scales.

With an increase in the detail to which paleoecological data can be extracted from individual specimens and bonebeds as a whole, it is essential to understand the context of fossil accumulations. Where paleosols contain fossil remains, they can be used to infer the method in which fossils were incorporated into the stratigraphic record. The relative degree of paleosol development at or near an archaeological or paleontological site can also provide information on the state of artifact and bone preservation (Holliday 2004). Understanding the depositional and post-depositional context of fossil and archeological sites is a key component to unraveling the taphonomy of the remains to be studied, the evolution of landscapes between depositional events, and can inform researchers on potential biases of preservation for a given locality.

Measuring and analyzing multiple paleosol profiles along a particular buried or exhumed paleo-CZ or paleocatena is a popular approach for reconstructing the paleoecosystem associated with an archaeological site. A catena study assesses the variability of soils across a landscape caused by changes in topography due to changes in hydrology (Morrison 1967; Birkeland 1999; Holliday 2004). Paleolandscapes can be studied at different scales depending on the research question being addressed. At the local scale, the relationship between depositional processes, weathering styles, drainage, and artifact accumulation can be examined. On the basin-scale, amalgamated strata representing a multitude of paleoenvironments can be used to understand long-term changes in climate and subsidence (Kraus 1999; Atchley et al. 2004; Cleveland et al. 2007, 2008; Aziz et al. 2008; Nordt et al. 2012; Atchley et al. 2013; Nordt and Driese 2013). An example of this method at Olduvai Gorge in Tanzania, where correlative surfaces buried by volcanic ashes have been used to reconstruct aspects of paleo-CZs from paleocatena successions (see description in Applications; Ashley and Driese 2000; Sikes and Ashley 2007; Ashley et al. 2014; Beverly et al. 2014, 2015a; Driese and Ashley 2016).

Biases and Shortcomings

In the spirit of Vance Holliday’s (2004) insightful review of paleosols as paleoenvironmental indicators, we also quote Valentine and Dalrymple (1976, p. 218), “Although soil science is under great pressure to furnish environmental evidence, it is debatable whether we understand the interaction of the soil-forming processes with the site and environmental factors well enough yet to make confident extrapolations.” This point alludes to the fact that many researchers make interpretive leaps to paleoenvironmental reconstructions without detailed understanding of the soil-forming and post-burial (diagenetic) processes that may have altered the paleosol system. This point should not deter future paleosol work. Rather, with a detailed analysis and an understanding of the biases discussed here, paleosol research using multiple proxies can yield a more nuanced interpretation of the paleoenvironment.

Disequilibrium with Climate or Environment

Soils are the product of local conditions, which may be considered an advantage in some paleoecological studies using paleosols; however, soils may not reflect the regional conditions due to the local hydrology, which may mimic changes in climate (e.g., Driese and Ashley 2016). It has been shown that multiple pedogenic pathways can yield the same soil property (Holliday 2004). For example, lateral changes in the development of a subsoil argillic (Bt) horizon from clay illuviation may be due to drainage or the duration of soil formation. This phenomenon, known as equifinality, suggests that in some cases a single paleoecological interpretation may not be possible when relying solely on the paleosol record. Thus, it may be difficult to extrapolate environmental and climate reconstructions beyond the site or pedon (Holliday 2004), but if correlation using precise dating methods or correlative tephras is possible, this issue can be avoided by studying the site as a paleocatena or paleo-CZ. By looking at the variations in a paleocatena across the whole landscape rather than focusing solely on the archaeological or fossil site, diachronous soil development can be identified and help researchers better interpret the paleoecology from paleosols. With this detailed method, the changes in local hydrology will be understood within the regional context, and paleoenvironmental and paleoclimatic interpretations can be made.

Pedogenesis on any substrate is a time-transgressive process and may not reach steady state with respect to climate and other environmental factors for upwards of 10³–10⁴ years (Yaalon 1971; Huggett 1998, and references therein). Therefore, less than centennial-scale changes may not be detectable in the soil or paleosol record and may reflect a sum of the environmental changes over the course of formation (Holliday 2004; Richter and Yaalon 2012). The relatively long durations of time required for soils to equilibrate with environmental conditions stands in stark contrast to the time required for the accumulation of artifacts or fossils, which can be instantaneous. Thus, a disparity of time resolution exists between the presence or burial of fossil remains and our ability to reconstruct the environment from which they are derived. This may result in disagreement between results of pedogenic and paleoecological or archeological proxies for environments. Young soils can have morphological and geochemical properties present in other (often drier) climatic settings. For example, excess base cations in immature soils will yield artificially low MAP and MAT estimates using climate proxies, resulting in “mock-aridity” or “mock-frigidity.” This should be kept in mind when interpreting climate change in stratigraphic sequences using paleosols of varying maturity (Stinchcomb et al. 2016).

Misuse of Proxies

Many paleosol proxies for climate have inherent limitations and are therefore subject to misapplication. For example, the depth to Bk and By proxies (Retallack 2005; Retallack and Huang 2010) predicts MAP based on the depth from the surface of a soil to the zone of nodular carbonate precipitation. As mentioned previously, many paleosols in the rock record are truncated by overlying fluvial deposits such that A horizons are a rarity in alluvial paleosols (Fig. 9.2C; Myers et al. 2014). This limits the depth to carbonate proxy to relatively few paleosols and must be applied with justification that A horizons were present or that the MAP value is an under estimation (Nordt et al. 2006; Prochnow et al. 2006). Early applications of the depth to carbonate proxy were performed without assessment of the origin of the carbonate, presence of surface horizons , or taxonomic interpretation of paleosols (e.g., Retallack 1997; Sheldon and Retallack 2004). It is now recognized that depth to carbonate in Vertisols follows a unique function (Nordt et al. 2006; Prochnow et al. 2006) that deviates from the more universal equation of Retallack (2005), which should caution researchers in application of such proxies in the absence of detailed paleosol description. Bulk geochemical proxies for MAP and MAT can be easily misused as well. Most paleosol proxies for climate are limited to specific soil orders or depositional conditions (Table 9.9). Misclassification of paleosols by horizons or soil order can therefore result in misapplication of climofunctions.

Perhaps a more widespread abuse of paleosols in climate studies is that very few researchers verify that the discrete geochemical values of paleosol samples fall within the range of values used to generate climate proxies. Because the relationship between a geochemical index and climate (e.g., CIA-K and MAP) is derived through a training data set, any applications of the regression equation must be confined to data that are similar to the training data . For example, unusually high concentrations of MgO (up to 17%) are present in paleosols at Olduvai Gorge because the parent material is an authigenic lacustrine Mg-rich smectite (Beverly et al. 2014). This is significantly more MgO than the dataset on which the CALMAG precipitation proxy was developed, which is only applicable in paleosols with <3% MgO (Nordt and Driese 2010a, b). This does not mean that these proxies are ineffective at predicting precipitation in diverse settings, but that care must be taken to correctly apply the proxies.

Stable isotopes of pedogenic carbonates are also commonly used as paleoecological indicators, but an understanding of soils and soil formation is necessary to apply these proxies. Researchers should interpret the presence of carbonate nodules or pedogenic carbonates in general with caution as several studies have shown that these features can be influenced by soil hydraulic properties (porosity, permeability), texture, clay content, and slope position (Holliday 2004). Also, the identification of carbonate in a paleosol does not mean that it is pedogenic in origin. For example, carbonate concretions are commonly found at Olduvai Gorge up to 20 or 30 cm in diameter and are associated with post-depositional groundwater movement through some paleosols (Fig. 9.2N), as indicated by their large size, irregular morphology, and sparry texture.

Reworking and Inheritance

Paleosols are components of sedimentary systems and are therefore subject to processes of erosion and deposition. The parent material of a soil reflects regional geology, which can include soils actively forming under variegated environmental (e.g., cl, o, r, p, t) conditions, bedrock of various compositions and windblown sediment. Any material removed from these reservoirs can be deposited on or incorporated into a soil. Geochemical proxies for climate (e.g., CIA-K) and vegetation (e.g., phytoliths, δ¹³C of organic matter) can be altered or biased if any constituent is included from reworked sediment.

Pedotransfer functions for climate that rely on paleosol bulk geochemistry will provide underestimates if sediment is added through any means of deposition. Geochemical climofunctions should not be applied to cumulative paleosols (Sheldon et al. 2002; Sheldon and Tabor 2009), which can form in eolian, fluvial, or mixed-source environments. Nordt and Driese (2010a) note that the smectite on which most Vertisols form is inherited from the surrounding catchment as alluvial clays rather than being authigenically manufactured in the soil profile. Other reworked pedogenic minerals, such as kaolinite or hematite, can make relatively immature paleosols appear to be mature, or can influence the color of a paleosol. Cross-referencing mineralogy with field and micromorphologic observations can mitigate misapplication or misinterpretation of paleoclimate in such settings.

Reworked organic matter and pedogenic carbonates are also a potential problem in alluvial deposits. Because paleosols typically have very low organic C concentrations (Wynn 2007), the relative abundance of in situ versus inherited organic matter can be difficult to discern. Rounded pedogenic carbonates found with coarser grains at the base of alluvial deposits are clear evidence for reworking (Lukens et al. 2017b). Comparison between SOM δ¹³C , pedogenic carbonate δ¹³C, and phytoliths has been a useful tool to reveal biases in the pedogenic carbonate vegetation signal (Cotton et al. 2012; Chen et al. 2015; Garrett et al. 2015). The difference between the δ¹³C of co-occurring organic matter and pedogenic carbonate is 14–17‰ in modern systems (Cerling and Quade 1993) but can potentially shift beyond that range given changes in atmospheric pCO₂ in the past (e.g., Mora et al. 1996) or isotopic fractionation of SOM during decomposition (Wynn 2007).

Diagenesis

Any alteration of paleosols after burial is referred to as diagenesis , and since the study of paleosols inherently involves analyzing soils that have left the cl, o, r, p, t conditions in which they were formed, one must carefully differentiate physical and geochemical features formed during pedogenesis versus those formed through diagenesis. Lithification is perhaps the most readily recognizable product of diagenesis and involves compaction and cementation. Compaction reduces pore volume, which may sufficiently inhibit fluid flow in clay-rich paleosols , but is unlikely to prevent groundwater transmission through paleosols with sandier textures. Compaction can also lead to deformation of mineral grains and reduction on primary porosity (Pettijohn et al. 1987).

Post burial cementation – not to be confused with pedogenic calcretes or fragipans – occurs in groundwater and plugs void space with a range of cementing agents. Calcite, silica, iron oxides, and clays are the most common cements in sedimentary rocks and are recognizable due to their occurrence within primary voids (e.g., root traces), along features of textural contrast (e.g., burrows infilled with sediment of an overlying deposit), or as overgrowths on mineral grains and pedofeatures (Stoops 2003). Calcareous paleosols typically show a mix of carbonate morphologies due to both pedogenic and diagenetic remobilization of primary pedogenic carbonates. The current rule of thumb in sampling pedogenic carbonate is to target micrite and avoid any material with sparry or megacrystalline texture (Sheldon and Tabor 2009; Tabor and Meyers 2015), though soils formed on coarse-grained sediments can contain pedogenic carbonates with sparry or microspar texture (Weider and Yaalon 1982). Any researcher attempting to study pedogenic carbonate for paleovegetation or paleoclimate reconstruction should first analyze carbonates in thin section and micro-drill targeted zones to avoid diagenetic overgrowths. Cathodoluminescence is a useful technique to identify histories of carbonate cementation (Barnaby and Rimstidt 1989; Machel et al. 1991; Michel et al. 2013) and should be considered before isotopic sampling is performed.

Diagenetic mineralization can alter the color of a paleosol to reflect the color of cementing agents or as a result of the dissolution of unstable minerals. Burial reddening occurs due to hematite cementation in oxidizing groundwater (Retallack 2001), whereas calcite cements can impart a white to pale color. Decomposition of organic matter in paleo-surface horizons will remove the dark colors evident in modern O and A horizons, leaving more uniform coloration to preserved paleosol profiles. If organic matter is decomposed by anaerobes after the paleosol is buried below the water table, blue-green drab halos will form around individual root traces or laterally along paleo-root mats (Retallack 2001). These colors can be differentiated from gleying due to saturation by identification of pedogenic features indicative of well-drained conditions, such as ped structure, translocation of clay or Fe, and precipitation of pedogenic calcite at depth.

Paleosols flooded by lacustrine or marine waters can exhibit a drab, reduced zone near their upper boundary (e.g., Driese and Ober 2005). This process is called pseudogley because it imparts gleyed colors typically observed in saturated soils. Pedogenic features indicative of episaturation should be present if gleying were due to periodic standing water on the soil surface, such as lateral-branching root traces, Fe-redox halos around roots and voids, pedogenic siderite , pyrite or jarosite, or accumulations of organic matter. Pseudogley has also been attributed to flushing of alkaline groundwater through buried paleosols (Pimentel et al. 1996).

Potassium is added during diagenetic conversion of smectite to illite (Fedo et al. 1995), which is the primary reason for excluding K₂O from pedotransfer functions and climate proxies (Harnois 1988; Maynard 1992). However, the physical pedogenic features indicative of smectite-rich Vertisols – including slickensides, sepic-plasmic clay fabrics and wedge peds – will remain after illitization has taken place (e.g., Driese and Foreman 1992). Current practice in the paleopedology community is to avoid using bulk geochemical data for climofunctions or pedotransfer functions if paleosols have experienced illitization or significant diagenetic mineralization (Sheldon and Tabor 2009; Driese et al. 2007; Medaris et al. 2017).

Pedogenic Mixing of Artifacts or Fossils

Pedogenic mixing processes can obscure discrete artifact layers and widen the depth range (Ferring 1992; Waters 1992; Cremeens et al. 1998). This can occur with aggressive bioturbation by soil fauna, through pedogenic processes such as cryoturbation from freeze-thaw cycles or due to vertic processes causing movement of artifacts or bones. Vertisol movement can also break and polish artifacts (Quade et al. 2004). However, even in Vertisols, which have considerable internal mobility of mass due to shrink-swell processes, artifacts can be placed in relative stratigraphic order with careful study (Waters et al. 2011).

Applications

The paleopedological techniques discussed in this chapter have been applied throughout geologic time, including diverse settings such as Neoarchean paleoweathering surfaces (Driese et al. 2011), late Ordovician and Silurian Appalachian Basin paleosols (Driese and Foreman 1992; Driese et al. 1992), Triassic paleosols at Petrified Forest, Arizona (Cleveland et al. 2007, 2008; Trendell et al. 2013a, b), paleosols spanning the Paleocene-Eocene Thermal Maximum in the Bighorn Basin, Wyoming (Kraus 1997; Kraus and Brown 1988; Kraus and Hasiotis 2006; Aziz et al. 2008; Snell et al. 2013) and Pleistocene-Holocene paleosols in eastern North America (Driese et al. 2005). Here we present examples from East Africa where paleosols have been used to understand the paleoecology at paleontological and archaeological sites throughout the Miocene and Quaternary.

There is a long history of paleosol research in East Africa that began with stable isotopes of C and O from pedogenic carbonates in the Turkana Basin (Cerling et al. 1988) and Olduvai Gorge (Cerling and Hay 1986). These extensive, well-dated paleosol records showed that climate became warmer and drier over the past 2 Ma (Cerling and Hay 1986; Cerling et al. 1988; Levin et al. 2011). This research ushered in a wave of studies focusing on stable isotopes of pedogenic carbonates to understand the paleo-vegetation and paleo-rainfall history at archaeological sites (Cerling et al. 1988; Wynn and Feibel 1995; Sikes et al. 1999; Wynn 2000; Levin et al. 2004; Quade et al. 2004; Wynn 2004; Lepre et al. 2007; Quinn et al. 2007; Campisano and Feibel 2008; White et al. 2009; WoldeGabriel et al. 2009; Cerling et al. 2010; Passey et al. 2010; Cerling et al. 2011; Levin et al. 2011; Quinn et al. 2013; Garrett et al. 2015; Lüdecke et al. 2016). For summaries of the use of stable isotopes in paleoenvironmental reconstructions at hominin sites see reviews by Cerling (2014) and Levin (2015).

This focus on stable isotopes of pedogenic carbonate has yielded long-term records during the Pleistocene for comparison with marine records but has also ignored the wealth of paleoclimatic and paleoenvironmental information that can be gained from the breadth of methods discussed in this chapter. Much remains to be discovered regarding the historical and process-based details of how hominins interacted with their surrounding ecosystem (National Research Council 2010). In this section, we focus paleosol applications at Olduvai Gorge to demonstrate the process and utility of paleosol research (Fig. 9.5). Other similar studies have recently been completed at significant fossil sites in western Kenya at Rusinga Island (Michel et al. 2014) and Karungu (Beverly et al. 2015a; Driese et al. 2016; Lukens et al. 2017b).

Fig. 9.5

Paleogeographic map of Olduvai Gorge modified from Ashley et al. (2010) and Beverly et al. (2014) showing paleo-Lake Olduvai and the locations of soil-forming environments. Cross-sections A and B are shown in Fig. 9.6

Olduvai Gorge is a well-dated archeological site that has been studied for over 50 years, but only recently have the paleosols been recognized for their paleoenvironmental potential (Ashley and Driese 2000; Hover and Ashley 2003; Sikes and Ashley 2007; Magill et al. 2013a, b; Ashley et al. 2014; Beverly et al. 2014; Driese and Ashley 2016). Using well-dated and laterally correlative tephras, Olduvai Gorge can be reconstructed as a series of paleocatenas or paleo-CZs at multiple scales of detail (Fig. 9.6; Ashley et al. 2014; Beverly et al. 2014).

Fig. 9.6

Reconstruction of a paleocatena or paleo-Critical Zone from Olduvai Gorge at two different scales reproduced from Beverly et al. (2014). The locations of cross-sections A and B are shown in Fig. 9.4. A. The local scale reconstructs a paleocatena along the margin of paleo-Lake Olduvai using field description of paleosols, micromorphology, and mass-balance of bulk geochemistry where very minor differences in topography and distance from the spring affected pedogenesis . B. The basin scale reconstruction is a summary of a decade of paleosol research and combines the use of bulk geochemistry, stable isotopes , mineralogy, and detailed field descriptions

Near the margin of paleo-Lake Olduvai a series of time-equivalent, stacked paleosols were identified and correlated using field relationships and tephrostratigraphy. Four trenches were described by horizon and classified as paleo-Vertisols and then sampled every 10 cm for bulk geochemistry. Oriented samples for micromorphology were also collected. The combination of micromorphology and bulk geochemistry proved to be a powerful tool in describing a paleocatena where minor differences in paleotopography due to small faults and distance from a fault-controlled spring affected the amount and type of pedogenesis . Close to the Zinj fault and a related spring on the downthrown fault block, stacked paleosols developed (Figs. 9.3D and 9.6A), but further away cumulative soils developed (Fig. 9.6A). Using micromorphology, the degree of pedogenesis could be assessed, and with UVf, the amount of organic matter could be visually estimated. Closer to the spring, weakly developed b-fabrics (monostriated), Fe–Mn filled rhizoliths, and high abundances of organic matter were observed. Pedogenesis was weaker due to a higher water table that did not allow for movement of water through the soil (Fig. 9.6A). Further from the spring, organic matter abundance decreased, but b-fabrics were better developed (parallel striated, cross-striated, and granostriated), volcaniclastic grains were weathered and pitted, and Fe–Mn and zeolites formed hypocoatings along ped and pore boundaries. When combined with the phytolith and pollen record from these paleosols (Barboni et al. 2010), a reconstruction of the paleocatena is complete with a groundwater forest closest to the spring and a more open environment with palms and grasses further from the spring (Fig. 9.6A).

Because the carbonates in these paleo-Vertisols were interpreted to be diagenetic precipitates based on their large, irregular size and sparry crystalline texture, bulk geochemical methods were utilized to reconstruct paleoclimatic conditions. An example of mass-balance from the bulk geochemistry is presented here using Na, Al, Si, and K, which are typically associated with clay translocation , but because this paleo-Vertisol formed on Mg-rich smectites, Mg is also included (Fig. 9.7A–B). Results of mass-balance calculations indicate that greater translocation occurs up-section, which is best illustrated with a box and whisker plot using Na as an example (Fig. 9.7C). The increased translocation overlaps with a precession cycle that has been previously identified in uppermost Bed I using stable isotopes , fauna, flora and lithostratigraphy (Cerling and Hay 1986; Ashley 2007; Barboni et al. 2010; Beverly et al. 2014, and references therein).

Fig. 9.7

Application of mass-balance to Olduvai paleosols. Reproduced from Beverly et al. (2014). A and B, Time-equivalent stacked paleosols near the lake margin of paleo-Lake Olduvai that can be correlated using field relationships and tephrostratigraphy. This example shows elements grouped by association with clay or zeolite accumulation. Ti is held constant as the immobile element (Tau Ti) and positive values indicate a net gain and negative values indicate a net loss relative to the parent material . Multiply by 100 to convert to percent gains or losses. C, The box and whisker plot of Tau – Na using Ti as the immobile element for each paleosol indicates that the amount of translocation increases up section

These methods can also be applied to a basin-wide reconstruction of the paleocatena or paleo-CZ (Figs. 9.5 and 9.6B). By analyzing these paleosols in a more holistic method that includes field relationships and careful description of paleosol features such as ped structure or redoximorphic features, stable isotopes , micromorphology, bulk geochemistry, and grain size (Table 9.11), a detailed reconstruction of the paleoenvironment in which hominins were living can be created (Fig. 9.6B; Ashley et al. 2014). Differences in parent material and topography have a great effect on paleosol development at Olduvai Gorge irrespective of climate. Paleo-Aridisols formed in the quartzo-feldspathic sediment on the fluvial plain to the west of paleo-Lake Olduvai (Fig. 9.2A–B) in contrast to the paleo-Vertisols formed in the Mg-rich smectitic clay along the lake margin (Fig. 9.2D–F). The topography of the basin has a great effect on the water table, and further from the lake, soil drainage increased and soils were generally redder in color due to oxidation of Fe (Ashley and Driese 2000; Sikes and Ashley 2007; Ashley et al. 2014). This example illustrates profoundly variegated soils that result from variations in climate, organisms, relief, parent material , and time (cl, o, r, p, t). Without careful study and correlation using tephrostratigraphy, the changes in paleosol types could easily be attributed to changes in one specific soil-forming factor such as climate – the paleo-Vertisol high in OM would have formed a wetter climate and carbonate-rich paleo-Aridisols would have formed in a drier climate. In actuality, these paleosols formed coevally on the landscape and the changes in relief and parent material were predominant controls on soil formation. With careful study, climate can be teased out of this complex signal, but climate cannot be the assumed cause of paleosol heterogeneity on the landscape (Figs. 9.5, 9.6A–B; Ashley and Driese 2000; Sikes and Ashley 2007; Ashley et al. 2014; Beverly et al. 2014).

Table 9.11 Summary of Olduvai paleosol characteristics modified from Ashley et al. (2014)

Future Prospects

The addition and refinement of paleosol proxies for climate and environment will advance paleopedology -based paleoecological reconstructions in the near future. The recent development of instrumentation capable measuring of clumped isotopes (Δ₄₇) in paleosol carbonates has allowed for great advances in reconstructing paleotemperature using paleosols. Ongoing research suggests that clumped isotopes may be seasonally biased and sensitive to soil hydroclimate (Passey et al. 2010; Gallagher and Sheldon 2016), and presents opportunities for future research. In addition, the development of the capability to measure ¹⁷O in addition to ¹⁶O and ¹⁸O presents new and interesting prospects in understanding the relationship between evaporation and precipitation using triple oxygen isotopes (Passey et al. 2014). New microanalytical techniques on pedogenic carbonate using U-series dating and laser ablation ICP-MS and ion microprobe to measure variations in δ¹⁸O and δ¹³C are also pushing forward the field of paleopedology (Oerter et al. 2016).

Larger modern soil data sets with more quantitative soil attributes will yield new models for predicting both environmental factors and soil properties. This may improve both climate prediction and the ability to more rigorously classify paleosols using Soil Taxonomy (Nordt and Driese 2013). Larger datasets may also provide a means for developing multiple models designed for specific soil characteristics (e.g., soil order, drainage, parent material ). Using a data-driven, multivariate statistical approach applied to a large soil dataset, Stinchcomb et al. (2016) showed that both MAP and MAT can be predicted simultaneously for a diverse range of soils forming across a large range of climatic conditions. This is especially useful when the paleopedologist has little environmental data at the outset of a project. Greater effort in estimating fundamental soil properties also has a bright future in the field of paleopedology . Soil pH and Eh proxies were used to construct a “paleo-Pourbaix diagram” used in the assessment of soil nutrient cycling and ecosystem response from Late Cretaceous and early Paleocene paleosols in the Western Interior of North America (Nordt et al. 2011). Similar efforts from East African and other locations could yield new insight into soil-plant-water interactions accompanying evolution.

Key Terms

• Catena – A series of closely-related soils that manifest unique properties due to lateral changes in soil-forming conditions that vary as a function of slope. The catena concept has been applied to paleosols as their position relates to valley bottom features such as springs and channels (see pedofacies).

• Critical Zone (CZ) – Near-surface environment in which complex interactions involving rock, soil , water, air, and living organisms regulate the natural habitat and determine the availability of life-sustaining resources (National Research Council 2001).

• Edaphic – Influenced, produced, or related to soil , typically in reference to mineralogical or chemical characteristics that affect biota.

• Eluviation – The transport and removal of soil solutes or material (mineral or organic) from an upper soil horizon . The material is commonly clay to silt size or dissolved in pore water.

• Illuviation – The introduction of soil solutes or material (mineral or organic) into a soil horizon from an overlying horizon. The material is commonly clay to silt size.

• Lithologic discontinuity – The contact between two different sediment types or genetically unrelated sediments that is usually marked by a distinct surface. When present, time may be unrepresented in the stratigraphic record.

• Master horizon – A dominant soil horizon within the soil profile. Master horizons include: O, A, E, B, C, R, and K. The B and C master horizons are the most common in the sedimentary record. Master horizons also may have subordinate horizon modifiers describing secondary characteristics within that soil horizon.

• Paleo-Critical Zone (paleo-CZ) – The deep-time equivalent of a modern CZ that is primarily studied using proxies for reconstructing past biogeochemical cycles related to environmental factors. The paleo-CZ is also known as a Deep Time Critical Zone (DTCZ). The DTCZ concept was originally confined to pre-Quaternary rocks and sediments (Nordt and Driese 2013). The concept is extended to the Quaternary (<2.6 Ma) in this chapter.

• Paleoecology – The study of animal and/or plant interactions within ancient environments. Studies may incorporate direct fossil evidence or rely on proxy methods.

• Paleopedology – The study of paleosols.

• Paleosol – A soil that formed on a landscape of the past or under soil-forming conditions no longer operating at a given locality. Paleosols can occur as buried, exhumed, lithified, and/or relict soils. Paleosols are also known as fossil soils.

• Parent material – The material (sediment or rock) from which a soil forms.

• Pedogenic – Adjective used for soil, soil-forming processes, or products of soil formation (e.g., pedogenic minerals)

• Pedology – A subdiscipline of soil science that emphasizes the study of soil formation, especially as it relates to environmental factors (e.g., climate, organisms, relief, parent material , and time).

• Soil – a depth profile consisting of distinct layers that result from the physical, chemical, and biological alteration (i.e., weathering) of pre-existing parent materials into a more stable form.

• Soil facies or pedofacies – lateral variation in soil development caused by variation in environmental factors of soil formation (i.e., topography).

• Soil horizon – A layer within the soil profile that is different from the overlying and underlying layers, usually based on physical features such as color and texture.

• Surface horizon – The zone within a soil profile dominated by organic matter inputs from the vegetative cover (e.g., A horizon).

• Subordinate horizon indicators – Lower-case letters that are appended to master horizon designations that add distinction to the master horizon (see Table 9.2 for a list of subordinate horizon indicators).

• Subsurface horizon – Also called the mineral soil , it is a zone where chemical reactions and physical actions cause the mobilization of ions and fine-grained particles, including clays, metals, and organo-metallic complexes (e.g., B horizon).

• Translocation – The movement (usually downward) of chemical and physical constituents through the percolation of water in pores or via biotic agents.

Key References

• 1. Soils in Archaeological Research (Holliday 2004) is devoted to soil and paleosol investigations as they relate to the archaeological record. This textbook is geared towards Quaternary paleopedologists and geoarchaeologists and has a chapter devoted to soils in paleoenvironmental reconstructions.

• 2. The Nature and Properties of Soils (Brady and Weil 2008) is a comprehensive introduction to modern soil science with an emphasis on chemical, physical, and mineralogical characterization of weathering as it pertains to the modern U.S. Soil Survey program.

• 3. Soils and Geomorphology (Birkeland 1999) is a well-known textbook on the dynamics of soil formation related to the five environmental factors: climate, organisms, relief, parent material , and time. This book is frequently used by modern and paleopedologists.

• 4. Guidelines for analysis and description of soil and regolith thin sections (Stoops 2003) is widely available text that serves as a step-by-step guide for analyzing and describing soil thin-sections and includes examples and a CD with images. This book is partially based on Bullock et al. (1985), which is more difficult to find.

• 5. Interpretation of Micromorphological Features of Soils and Regoliths (Stoops et al. 2010) is a textbook with excellent color images illustrating features that can be identified in soils and paleosols.

• 6. Stable isotope evidence for hominin environments in Africa (Cerling 2014) is a chapter summarizing how stable isotopes can be used to reconstruct hominin environments.

• 7. Environment and Climate of Early Human Evolution (Levin 2015) is another summary of the use of stable isotopes to reconstruct hominin paleoenvironments.

• 8. Quantitative paleoenvironmental and paleoclimatic reconstruction using paleosols (Sheldon and Tabor 2009) is an extensive summary of the state of paleosol research and is good place to start for those interested in utilizing paleosols for paleoenvironmental and paleoclimatic research.

• 9. Paleosols as Indicators of Paleoenvironment and Paleoclimate (Tabor and Myers 2015) is an update to the Sheldon and Tabor (2009) summary of paleosol research.

• 10. Revisitation of Concepts in Paleopedology: Transactions of the Second International Symposium on Paleopedology (Follmer 1998) is a collection of abstracts, essays and reports by various authors on numerous facets of paleosols, following the first paleopedology symposium (Yaalon 1971).

• 11. Weathering, Soils & Paleosols (Martini and Chesworth 1992) is an edited volume that is a blend of pedology, classic geology (especially mineralogy and geochemistry) and geomorphology that discusses present and past weathering processes. Although this book offers an intriguing and unique view of paleopedology , it contains little content on paleosols as paleoecological indicators.

• 12. A Colour Guide to Paleosols (Retallack 1997) is a hard-to-find introductory book on paleopedology that contains remarkable paleosol images and accessible text on the basics of the field, including paleosol identification , description and interpretation.

• 13. Soils of the Past: An introduction to paleopedology (Retallack 2001) is a paleopedology textbook that introduces concepts and techniques in the field and provides a chapter-by-chapter overview of paleosols recorded through Earth history and their paleoenvironmental importance.