groundwater-fed wetland sediments and...

GROUNDWATER-FED WETLAND SEDIMENTS AND PALEOSOLS: IT’S ALL

ABOUT WATER TABLE

GAIL M. ASHLEY

Department of Earth and Planetary Sciences, Rutgers University, Piscataway, New Jersey 08854, USA

e-mail: [email protected]

DANIEL M. DEOCAMPO

Department of Geosciences, Georgia State University, Atlanta, Georgia 30302, USA

JULIA KAHMANN-ROBINSON

Department of Geology, Baylor University, One Bear Place #97354, Waco, Texas 76798, USA

AND

STEVEN G. DRIESE

Department of Geology, Baylor University, One Bear Place #97354, Waco, TX 76798, USA

ABSTRACT: Wetlands are continental depositional environments and ecosystems that range between ephemerally wet to fully aquatic habitats, and, thus,

the character of a wetland soil is directly related to the position of the water table over seasonal and longer timescales. The sediment and paleosol records

of wetlands are products of a unique setting that can be both exposed to the atmosphere and water-saturated at the same time. Wetlands tend to occupy

low-gradient portions of the landscape in places where the phreatic zone is at least ephemerally exposed at the surface, and hydrophytic vegetation has an

opportunity to colonize. Groundwater-fed wetlands are an end member of a continuum of waterlogged environments and are associated with localized

groundwater discharge (GWD); e.g., springs and seeps that can sustain permanent saturation. Research has tended to follow one of two parallel tracks:

sedimentology or pedology. An objective of this paper is to bring these two separate lines of inquiry closer together. The signature of wetland

pedogenesis includes redoximorphic features, enhanced hydrolytic alteration or dissolution of soluble phases, and preservation of biotic indicators of

wetland habitats. Histosols (peats) and other hydric soils (indicated by gley color and reduced minerals like pyrite and siderite) are common in sites with a

permanently high water table and anaerobic conditions. Illuvial clays, in contrast, record episodes in which wetlands dry out and drainage improves

sufficiently for these features to form. A case study from Holocene-age Loboi Swamp, Kenya, illustrates the importance of integrating field observations

and laboratory analyses. Wetland conditions were observed through thin section micromorphology, mineralogy, bulk geochemistry, and macro- and

microfossils. The record of Loboi Swamp is characterized by the juxtaposition of features indicating episodes of soil saturation alternating with those

indicating desiccation. In order to extract the most information recorded in groundwater-fed wetlands, soils and sediments should be studied as part of the

larger spatial and climatic frameworks in which they occur.

KEY WORDS: wetland, groundwater, paleosol, micromorphology, hydric soils

INTRODUCTION

Wetlands are depositional environments and ecosystems that rangebetween ephemerally wet to fully aquatic continental habitats. Thereare numerous landscape contexts in which inland (nontidal) ephemeralwetlands develop in response to seasonal changes in water budget.However, perennial wetlands require a sustained water source (e.g.,groundwater discharge locations determined by topography orgeology) coupled with an impermeable or poorly drained substrate.The sediment and paleosol records of groundwater-fed wetlands areproducts of a unique setting that can be both subaerial (exposed to theatmosphere) and water-saturated at the same time and because of thisdichotomy are often misinterpreted or not recognized for what they are.In wetlands, surficial sediment may be exposed to the atmospherewhile the water table is at the surface or mere millimeters below. Thismay produce contradictory indicators in associated sediments andpaleosols, potentially leading to misinterpretation.

Wetlands tend to occupy depressions and low-gradient portions ofthe landscape in places where the phreatic zone is at least ephemerallyexposed at the surface, and hydrophytic vegetation, such as sedges

(e.g., Carex, Cyperus), cattails (Typha), mosses (Sphagnum), and

others, has an opportunity to colonize (Fig. 1). Such conditions have

been documented on hillslopes, fluvial floodplains, and exposed lake

flats following lacustrine regression. They have traditionally been

studied in isolation rather than as just one component of a larger

landscape. Wetlands are an important component of the critical zone,

‘‘the heterogeneous, near surface environment in which complex

interactions involving rock, soil, water, air and living organisms

regulate the natural habitat and determine availability of life sustaining

resources’’ (NRC 2001, p. 2). Pedology and paleopedology are now

moving into research areas that view soils as part of the critical zone,

including the larger spatial and temporal contexts of hydrologic and

geochemical cycles (Wilding and Lin 2003, Lin et al. 2006, Sheldon

and Tabor 2009, Brantley et al. 2011) and paleoclimatic records

(Sheldon and Tabor 2009). Interpretation of wetlands in the geologic

record requires a broad spatial perspective based on sedimentary facies

associations (Deocampo 2002a). Paleosols from wetland settings may

retain paleoclimatic and paleoenvironmental records controlled by

depositional setting and hydrogeologic context. Thus, their interpre-

New Frontiers in Paleopedology and Terrestrial Paleoclimatology: Paleosols and Soil Surface Analog Systems j DOI: 10.2110/sepmsp.104.03SEPM Special Publication No. 104, Copyright � 2013SEPM (Society for Sedimentary Geology), ISBN 978-1-56576-322-7, p. 47–61.

This is an author e-print and is distributed by the authors of this article. Not for resale.

tation requires a broad temporal perspective, recognizing dynamicatmospheric, hydrospheric, and biospheric changes through time.

Because they tend to occupy low-gradient portions of the landscape,wetlands are sites of sediment accumulation and the gradual buildup ofan organic-rich record. However, because sedimentation rates inwetlands tend to be low, the sediments are typically subjected toextensive bioturbation by those very infauna and flora that generate theautochthonous sediment (Hasiotis 2000, Hasiotis et al. 2007). Wetlandresearch using a sedimentological approach includes grain size,magnetic susceptibility, clay mineralogy, ichnology, stable isotopesof both carbonate and organic matter, records of sulfur, nitrogen,phosphorous, etc., and biological remains (diatoms, pollen, plantmaterial) (Quade et al. 1995, Owen et al. 2009, Rucina et al. 2010). Inwaterlogged environments, geomicrobiological processes (e.g., micro-bial respiration) strongly control geochemical parameters such as redoxstatus, pH, and aqueous CO2 levels. These impact the thermodynamicstabilities of carbonate, silicate, and sulfide mineral systems (Baas-Becking et al. 1960, Deocampo and Ashley 1999).

Wetland sediments, and ultimately the soil and paleosol records ofthem, are the end products of the influences of climate, local geology(tectonics, source lithology, and topography), hydrology, and wetlandbiota (Fig. 2). The character of a wetland soil, specificallygroundwater-fed wetlands, is directly related to the position of thewater table over seasonal and longer timescales; the many factorscontrolling sedimentation create challenges in deciphering the records.Research on wetlands has tended to follow one of two parallel tracks:sedimentology or pedology. This paper attempts to bring these twoseparate lines of inquiry closer together. The goals are to (1) examinethe environmental and hydrological contexts of groundwater-fedwetland sediments and associated pedogenesis, (2) summarize theprocesses and resulting records, and (3) present a descriptive faciesmodel for recognizing groundwater-fed wetland paleosols in the rockrecord.

ENVIRONMENTAL CONTEXT

Geomorphology

Wetlands that are likely to leave a permanent paleosol record aresituated in areas where the substrate is at least periodically saturatedwith water. These are found at the base of hillslopes, in central portionsof basins (Weissmann et al., this volume), and in locations whereimpervious beds intersect the surface creating a perched water table, oralong fault traces (Armenteros et al. 1995, Quade et al. 1995, Driese etal. 2004). This environment is referred to as palustrine and includes arange of semipermanently to permanently inundated nontidal environ-ments (Fig. 1) (Mitsch and Gosselink 2000). Wetlands may bemaintained by river inflows (Tooth and McCarthy 2007), but surfacerunoff can be variable and insufficient to maintain a wetland. Incontrast, groundwater flow (which is protected from evaporation) ismore likely to provide a sustainable water supply during short- andlong-term climate fluctuations (Ashley et al. 2004, Driese et al. 2004).Sites that are wet enough to support hydrophytes and produce hydricsoils are usually fed by groundwater (Renaut 1993; Deocampo 2002a;Melchor et al. 2002; Liutkus and Ashley 2003; Ashley et al. 2010a,2010b).

Hydrology

At local scales, climate (i.e., absolute precipitation and temperature)is less important than hydrogeologic context and presence of aquifersin the area. Rainfall that is not evaporated, intercepted, or transpired byplants infiltrates into groundwater in recharge zones. It movesdownslope under the influence of gravity or hydraulic head (Fig. 3).Evapotranspiration of groundwater is typically negligible, and is thusbuffered against short-term changes in water budget. The water movesslowly (m/yr) through bedrock, and the local geology (structure and

FIG. 1.—The palustrine environment. The diagram depicts examples of inland wetland habitats on an undulating land surface. The difference in

position of the water table from saturated (6) to temporarily flooded (1) affects vegetation type and ultimately the nature of the wetland

(modified from Cowardin et al. 1979).

48 GAIL M. ASHLEY, DANIEL M. DEOCAMPO, JULIA KAHMANN-ROBINSON, AND STEVEN G. DRIESE


lithology) determines the path and the locations where flow dischargesonto surface (Fig. 3). Accumulation of sediments is greatest intopographic lows and the most likely location for soils to form and bepreserved.

Rivers and lake margins are typically the lowest areas on the

landscape, and, as in palustrine settings, the ‘‘flood state’’ is dictated bythe height of the water table (Fig. 1) (Cowardin et al. 1979, Mitsch andGosselink 2000). Thus, water levels may be (1) temporary, (2)seasonal, (3) semipermanent, (4) intermittent, or (5) permanent inriverine and lacustrine environments. However, the flood state ofsaturated (6) is much less common in rivers because of fluctuatingwater supply and the vagaries of surface runoff. Thus, the hydrologicconditions of groundwater-fed wetlands in palustrine settings are likelyto be the most consistent, creating a permanently saturated or floodedenvironment (van der Kamp and Hayashi 2009). Groundwater-fedwetlands, the focus of this paper, therefore represent an end member ofa continuum of wetland types in nature.

PROCESSES OF FORMATION

Sediments

Groundwater-fed wetlands accumulate both allochthonous andautochthonous sediment (Fig. 2). The allochthonous sediment cancome from colluvium washed in from higher ground, detritus blown infrom exposed sand or mud flats (Fig. 4A), and pyroclastic air fall involcanically active areas. Autochthonous sediment varies with thetypes of plants and animals present; the biogeochemical environmentof sediment and water determines the composition of the authigenicminerals that form in situ. This environment, in turn, is determined bythe position of the water table and the local water budget (particularlyevapotranspiration, which can increase solute concentrations).

FIG. 2.—Flow chart showing the major controls and feedbacks on sedimentation in groundwater-fed wetlands (modified from Deocampo 2002a).

FIG. 3.—Cartoon of the hydrogeology of groundwater-fed systems

showing a recharge region on a nearby topographic high and

groundwater flow path downslope dictated by the local geology

(bedding, folds, and faults). Aquifers allow the steady, sustained

supply of water needed to maintain groundwater-fed wetlands

(modified from Marshak 2004).

GROUNDWATER-FED WETLAND SEDIMENTS AND PALEOSOLS: IT’S ALL ABOUT WATER TABLE 49


Studies of modern and Holocene wetland deposits indicate that they

produce generally hydric paleosols composed of peat, localized tufas,

and organic-rich clay deposits that contain eolian-transported mineral

matter, macrophyte plant remains (e.g., roots, stems), pollen,

phytoliths, diatoms, charcoal, carbonate (calcite and siderite), and

manganese-rich nodules (Marchant and Taylor 1998, Deocampo

2002a, Wust and Bustin 2004, Rucina et al. 2010). Older deposits

retain little original organic matter, and the plant remains are often

silicified, but otherwise the record is similar (Liutkus and Ashley 2003,

Norstrom et al. 2009). Systematic analyses of biological records

(diatoms, ostracodes, pollen, and testate amoeba) document climate

change, as well as the quality of the water (total dissolved solids,

alkalinity, and oxygen levels) over time (Charman 2001, Grosjean

2001, Owen et al. 2004, Johnson et al. 2009).

Biota

The groundwater environment is biotically distinct from the surface

environment, and so the groundwater-fed wetlands lie at an ecotone

between surface and subterranean ecosystems; this is reflected by

biotic records (Quade et al. 1995, Reeves et al. 2007). The critical zone

sediment is home to a range of microbes including fungi, fauna

(protozoa to mammals), and flora (algae to trees). Although there is life

below the water table, it is less abundant than in the vadose zone(Hasiotis 2000, Retallack 2001, Hasiotis et al. 2007).

The records left by biota are biogeochemical (organic compoundsand isotope signatures) or physical; i.e., body fossils and ichnofossils(traces of animal behavior), as well as root traces. However, becausemost roots facilitate aerobic respiration, they usually do not penetrateinto the phreatric zone. In the waterlogged environment of wetlands,the roots tend to spread out laterally in order to ‘‘breathe’’ (Fig. 4B),but, if the water table fluctuates seasonally, vertical root traces can alsobe created during times of low water table. The resulting matrix ofvertical and horizontal root traces complicates the interpretation of thelong-term record (Coleman 1988, Retallack 2001).

The proportion of time that the wetland is completely submerged vs.time the water table fluctuates affects all pedogenic processes (physical,chemical, and biological). The water table is that level below which theground is saturated (phreatic zone). The redox state of the phreatic zoneclose to point sources of discharge is controlled by the redox state ofgroundwater in the discharging aquifer; hence, they may be reducing oroxidizing depending on the hydrogeologic context. Phreatic watersfarther from discharge points are more likely to be reducing due toorganic matter accumulation. The vadose zone is more uniformlyoxidizing and is churned continuously by organisms. This bioturbationenhances exchange of gases from the atmosphere, but the exchangedecreases downward (Fig. 5). Oxygen decreases and PCO2 increases (aby-product of microbial breakdown of organic matter) with depth.There is also an increase in relative humidity and soil moisture and aconcurrent decrease in biodiversity and biotic exchange, so there is agradual decrease in organic matter content and intensity of bioturbationwith depth.

Hasiotis developed a tiered classification of ichnofossils based onmoisture requirements and tolerances (Fig. 5) directly tied to theposition of the water table (Wallwork 1970, Hasiotis 1997, Hasiotis2000). The top tier consists of terraphilic (e.g., beetles, ants, termites,nematodes, etc.) traces constructed by upper-vadose-zone organismswith a low tolerance for water that thrive best in a well-drainedenvironment. The next tier down is hygrophilic traces created by bees,wasps, and some mollusks, etc., that occupy the lower part of thevadose zone and must have some water (particularly in the form of soilmoisture). The lowest tier consists of hydrophilic traces that areconstructed by organisms that are aquatic and live at or below the watertable, such as subterranean ostracodes, and some types of mollusks,crabs, and crayfish (Thorp 1949). When the water table rises to thesurface during seasonal or long-term positive changes in hydrologicbudget, then the tiers compress or disappear and hydrophilic traces arefound at the surface (Fig. 5).

Geochemical Environment

A wide variety of biogeochemical reactions can occur in wetlands,including both acid–base and oxidation–reduction reactions. Acid–base reactions in wetlands depend primarily on major solutecompositions of inflow waters. Major solute compositions in wetlandsare controlled principally by the geochemistry of local groundwater,which is in turn controlled by weathering reactions in the watershed. Inhydrologically closed wetland systems, where evapotranspiration is theonly route for water to exit the wetland, evaporative concentration canquickly supersaturate waters with respect to carbonates, resulting incalcite precipitation (Garrels and Mackenzie 1967, Jones andDeocampo 2003, Deocampo 2010). However, in settings where soil-respired CO2 is not efficiently degassed, CO2 can build up. Thissuppresses pH, prevents carbonate precipitation, and enhancespreservation of biogenic amorphous silica (Deocampo and Ashley1999). In settings where eolian input adds soluble material (e.g.,sodium carbonate) to wetlands, the resulting solutes can also stronglyaffect the geochemistry of the waters (Deocampo 2004).

FIG. 4.—A) Photo of windblown dust in transport toward a

groundwater-fed wetland, Gorigor Swamp, Ngorongoro Crater,

Tanzania. Eolian detritus and pyroclastic material are the dominant

detrital source lithologies to groundwater-fed wetlands. B) A

lowered water table exposes the root system of two fever trees

(Acacia xanthophloea) in a groundwater-fed wetland. Note the tree

roots spread laterally above the former high water table during

growth.



Oxidation–reduction reactions are also complex in wetland systems.The water table separates the oxygen-rich, relatively dynamic vadosezone, where mineral weathering and leaching take place, from thestagnant, anaerobic chemically reducing environment of the phreaticzone (Fig. 5). The geochemical conditions possible on Earth are plottedon a graph of Eh (in millivolts) vs. pH scale that expresses a range ofacidic to alkaline conditions (Fig. 6A) (Garrels and Christ 1965, Faure1991). However, the range of conditions in the natural environment ismuch more limited and clustered around conditions of neutral pH and 0Eh (Baas-Becking et al. 1960) (Fig 6A). Earlier sedimentologists hadapplied this geochemical framework to explain stable mineralassociations that formed under different conditions of aeration (Eh)and acidity (pH) in sediments and soils (Krumbein and Garrels 1952,Blatt et al. 1960). A version of this mineral plot is now used by bothpaleopedologists and sedimentologists to interpret paleoenvironmentalconditions (Retallack 2001, Prothero and Schwab 2004) (Fig 6B).

Several geochemical fences (e.g., limestone fence and sulfate–sulfide fence) segment Eh–pH space into component stability fields.Thus, the paleo-Eh and paleo-pH of a paleosol can be approximatedfrom mineral assemblages, assuming no significant postdepositionalalteration. Extent of diagenesis can be assessed with thin sections ofsoil micromorphology. One of the dominant pedogenic processes inwaterlogged soils is gleying, the reduction of iron (to Fe2þ) and Mn (tomanganese Mn2þ or Mnþ), which alters the sediment to a green–blue–gray color for iron and purple–black for manganese (Vepraskas 1994,2001). When exposed to air, the gley colors change to a mottled patternof reddish, yellow, or orange patches due to oxidation of iron andmanganese (Hurt et al. 1998).

FIELD AND LABORATORY INDICATORS

Wetland Sediments

The sedimentological record is studied using grain size, color,sedimentary structures, mineralogy, biological remains, and bulkgeochemistry. Thus, there are many similarities with the pedologicalapproach, but the sedimentary record is interpreted from bottom to top;i.e., the history of wetland sedimentation, whereas soils are usuallystudied from the top down. However, aggrading systems are complex

in that the end product is a result of the interplay among deposition,erosion, and the rate of pedogenesis (Kraus 1999).

Sedimentary Structures: Root traces, bioturbation by terraphilicand hygrophilic and hydrophilic organisms, and discontinuous laminaproduced by changes in grain size or mineralogy can be seenmacroscopically and in thin sections (Hasiotis 2000) (Fig. 5). X-rayscans reveal finer-scale features.

Mineralogy: As wetland sediments tend to be fine-grained,magnetic susceptibility scans are useful in tracking changes inmineralogy (Gell 2010). Magnetic susceptibility is the degree towhich a material can be magnetized in an external magnetic field. Itmeasures the preferred orientation, distribution, or shape of magneticminerals and is a proxy indicator for changes in composition. Claymineralogy analysis can help distinguish allochthonous (detrital) claysfrom autochthonous (authigenic) clays that are formed after deposition(Singer 1980, Hover and Ashley 2003).

Biological Remains: Physical remains of plants (pollen, seeds,phytoliths), as well as roots and leafs, provide an important record ofmacrovegetation and the ways in which it may have changed over time.Diatoms are photosynthetic algae that precipitate silica tests. Testateamoebae are single celled, with either agglutinated or carbonate tests.Ostracodes have been used extensively in analysis of lacustrinesediment, and they are equally valuable in wetland sediments (Schwalbet al. 2002). Microfossil records are valuable in reconstructing theaquatic geochemistry (alkalinity, pH, and dissolved oxygen, etc.) of thewetland (Charman 2001, Owen et al. 2004).

Biogeochemistry: Organisms living in the wetland produce organicmatter with distinctive biochemical compositions, and total bulk %nitrogen and % carbon in the sediments reflect the type of plants thatwere present (Olago et al. 2003). C/N ratios are used to determinerelative proportion of nonvascular plants (phytoplankton and algae)and terrestrial vascular plants (Das 2002). Carbon isotopes of organicmatter, d13C, and d15N also provide information on the type of plants(e.g., C3 or C4 plants) that were present (Ziegler 1995). The abundanceof sulfide minerals is a measure of redox and is produced in a reducingenvironment by the reduction of sulfates by sulfur-fixing bacteria, such

FIG. 5.—Schematic diagram of the near surface showing the relationship of types of trace-making (burrowing) organisms to the position of the

water table (WT). The vadose zone (above WT) is an area of intense physical and chemical activity; i.e., gas exchange with atmosphere and

movement of water up (capillary action and evaporation) and down under gravity. Oxygen levels and organic matter decrease downward.

Organisms that can tolerate or have an affinity to some moisture (terrafiles and hygrophiles) occupy the vadose zone, whereas water-loving

organisms (hydrophiles) thrive in the phreatic zone (modified from Hasiotis 2000).



as Desulfovibrio (Mitsch and Gosselink 2000). Phosphorous levels are

usually a reflection of the former trophic state of the wetland.

Wetland Paleosols

Horizonation: Soil profiles are integrated records of sedimento-

logical, chemical, and biological processes during pedogenesis. The

changes in color, composition, texture, and pedogenic features are

permanent records of these processes preserved in paleosols (Fig. 7).

The profiles of waterlogged soil of wetlands are generally not well

developed because of the suppression of most of the soil-forming

processes. They are best seen in fresh excavations (Fig. 7A) or cores

(Fig. 7B). Figure 7A shows a 1-m excavation into a Holocene wetland

(Loboi Swamp, Kenya) that revealed a buried Inceptisol overlain by the

modern wetland soil. Radiometrically dated seeds at the interface place

the sharp contact between the soils at 640 to 720 year BP (Ashley et al.

2004, Driese et al. 2004). The lower soil was interpreted to have

developed on a silty–clay floodplain. The A horizon is missing, and the

master B horizon consists of two horizons (Bw1b and Bw2b) based on

slight color differences. The wetland soil has a thin (1 cm) peaty O

horizon at the surface, underlain by the A horizon, which consists of

organic-rich silty clay. The A horizon also contains spongy Typha roots

FIG. 6.—Aqueous geochemistry and mineralogy associated with varying Eh and pH of water. A) The range of conditions possible; box delineates

the normal range of conditions in wetlands (Eh 0.2 to�0.4 and pH from 6 to 0) (Garrels and Christ 1965). B) Minerals that form in wetlands

under different acidic/alkaline or redox conditions. Important thresholds in Eh–Ph space are shown by the position of the organic matter fence,

the sulfate–sulfide fence, the Fe, Mn oxide–carbonate fence and the limestone fence (Krumbein and Garrels 1952).



up to 5 cm in diameter and redoximorphic features (black Fe/Mn-oxidecoatings), clay-lined root traces, and insect macropores.

Cores of wetlands can be meters in length, with the advantage ofrecovering long pristine records, but cores are also narrow (;5–7 cm),which results in limited sample size to represent a given depth. Figure7B shows a 53-cm-long silty–clay core of a groundwater-fed wetland.The cored sediment has a peaty surface layer (left), color change frombrown to gray–green downward (right), and multicolored (brown–green) laminae at the base. The color variations were likely producedby a history of fluctuating water table and associated change in redoxconditions.

Root Traces: Wetlands are some of the most carbon-productivecontinental environments, producing from 147 to 1800 g m2/yr(Bradbury and Grace 1983, Keddy 2000). In addition to organic matter,macro- and microfossils, root traces (Fig. 7C), and rhizoliths (Fig. 7D)are definitive records of the larger plants that are part of the wetlandecosystem. Roots generally branch downward, but they are anatom-ically simple and difficult to correlate with a specific plant (Stewart andRothwell 1993, Retallack 2001). Traces of coarse (.1.5 mm diameter)and fine (,1.5 mm diameter) roots provide pathways for groundwatermovement and translocation of clay and precipitation of hematite,limonite, and other Fe or Mn oxides that create a visible stain definingthe trace. Figure 7C of a paleo-Vertisol shows dispersed fine rootstructures and a concentration of traces that likely represent a fossilroot mat (Beverly 2011). Figure 7D shows carbonate root casts(rhizoliths), which are sediment- and/or cement-filled root molds thatare postdepositional; they are pedogenic, and early studies interpretedthem to form exclusively in the vadose zone (Klappa 1980). Recent

studies using stable isotope geochemistry (Liutkus et al. 2005) orbiological remains and ichnology (Owen et al. 2009) and paleopedol-ogy (Kraus and Hasiotis 2006) have identified ‘‘phreatic rhizoliths’’that form in wetlands or poorly drained soils and are interpreted torepresent a fluctuating water table, where some, if not all, carbonate isprecipitated by groundwater.

Color: Sediment color (value and chroma) has been used tointerpret the hydrologic history of wetlands. Low-chroma colors (,3;e.g., gray or bluish-gray) are sediments depleted of iron (Fe) andmanganese (Mn) and are typical of water logging (Stoops and Eswaran1985, Vepraskas and Guertal 1991, Retallack 1997). Green sedimentsrecord a reduced environment (Fe2þ and Mn2þ), and red colors(hematite, limonite, and Mn oxides) indicate oxidizing conditions (Fig.6B). Mottled red and gray colors are a record of fluctuations of redoxconditions (Kraus and Hasiotis 2006), but colors can be inherited oroverprinted, and color change can occur rapidly. Redox potentialmeasurements in a field-based study showed that an average of 21consecutive days of continuous saturation was sufficient for Fereduction to occur in soils (He et al. 2003).

Pedogenic Structures: Groundwater-fed wetland paleosols haverelatively simple profiles compared to those developed in otherenvironmental settings. An organic-rich (peaty) O horizon, which canbe centimeters to meters thick, occurs at the top (Fig. 7A). The Ahorizon, which is a mixture of minerals and partially decomposedorganic matter, can vary considerably in thickness. The underlying Bhorizon is a record of the fluctuating water table and is commonlymottled gray–red–yellow with a mix of gleyed and oxidized minerals.The B horizon may contain cutans and weakly developed peds.Hematite and goethite nodules have also been reported (Retallack2001, Kraus and Hasiotis 2006), and root traces and burrows occur(Fig. 5) (Hasiotis 2000). The bottom part of the profile consists ofsediments that are nearly permanently waterlogged, with littlebiological activity, and consequently much of the relict bedding maybe preserved. Organic matter is preserved, and pyrite and sideritenodules record the chemically reducing environment (Ho and Coleman1969, Driese et al. 2010).

Morphological indicators of soil wetness (coatings and nodules)have been used with some success to develop metrics that will helpestablish ‘‘how wet is wet’’ (Vepraskas and Guertal 1991). Theirliterature-based study reported on the presence of depletion coatingsthat form by movement of Fe-Mn out of the groundmass and intomacropore voids. These features were also noted by Stoops andEswaran (1985). A survey of the literature by Vepraskas and Guertel(1991) indicated that if 40 to 80% of the wetland sediment hasdepletion coatings, then the wetland is saturated with waterapproximately 50% of the year. More basic research needs to be doneon the kinetics of the system to better understand the link between theduration of saturation and change in states of minerals.

FACIES ASSOCIATIONS AND PALEOCATENAS

Sedimentary facies models, which generalize and idealize the three-dimensional geometric, lithologic, and biologic characteristics of adepositional environment on a landscape scale, have been in use fordecades (Reineck and Singh 1975, Walker 1984, Reading 1996).Initiated by needs of the petroleum industry, facies models are normsthat can be used as basis for interpretation of incomplete records andpredict what is in the subsurface. An excellent example of asedimentary facies model for groundwater-fed wetlands was developedwith data from Tanzania, from a number of modern examples on lakemargin flats (Fig. 8) (Deocampo 2002a). The model is at landscapescale (a few km2), and the groundwater source is recorded permanentlyby a freshwater carbonate tufa mound. Groundwater moves up a fault

FIG. 7.—Photos. A) A 1-m-deep excavation in Loboi Swamp reveals

two partially developed paleosols; the lower one developed on a

preexisting floodplain, and the upper one formed in the modern

swamp. A radiocarbon date on seeds indicates the modern swamp

began to develop circa 640 to 720 year BP (Ashley et al. 2004).

Very finely crystalline siderite was found in the sediments,

suggesting reducing/alkaline conditions (Fig. 6). B) A 53-cm ore

collected from a wetland reveals organic matter (peat) at the top and

gradual change in gray–green coloration toward the bottom, where

evidence of redox changes is evident (mottled red and gray). Core

top is to the left (Shilling 2012). C) Chunk of Pleistocene-age

Vertisol composed of small (;5 mm long) Ti-Mn-oxide-filled root

traces suggests they were formed in the vadose zone under

oxidizing conditions (Beverly 2011) (Fig. 6). D) Calcite rhizoliths

formed in and beneath a groundwater-fed wetland from ground-

water saturated in calcium. Up is to the top (Liutkus et al. 2005).



under hydraulic head and flows onto the surface. A positive hydrologicbudget of groundwater discharge vs. evaporation proximal to thesource supports a freshwater habitat consisting of wetlands and openwater, used in this case by hippos. Distal from the water source, thehydrologic budget is negative, and the environment consists of adrainage stream ending in terminal evaporative, alkaline mud flats. Thesedimentary record is a mosaic of drainage channel sands, biogenicamorphous silica (from wetland plants and diatoms), and evaporites,carbonates, and Mg-clays from the alkaline flats. The wetland mud flatsare intercalated with lake mud flats. Bioturbation on all scales ispervasive, including hippo trails that can be a few meters wide and ameter deep. This facies model developed in the modern landscape hasbeen successfully applied to Pleistocene-age deposits (Liutkus andAshley 2003, Owen et al. 2009).

The concepts of facies associations using uniformitarianism andWalther’s facies law have not been easy to apply to buried paleosols(i.e., paleocatenas) (Valentine and Dalrymple 1976). The difficultiesstem from the possibility that several formative factors such as thehydrologic and geomorphic processes combined with sedimentology(i.e., parent material) may produce deposits that look like buried soils(Holliday 2004). As soils develop on the surface, it is important to beable to trace variation in the unit across a fossil landscape in order toverify a paleocatena (Valentine and Dalrymple 1976). Diagenesis mayalso alter some of the six pedogenic features or processes that are keyto identifying a paleosol: organic matter content, horizonation, redoxconditions, in situ mineral alteration, illuviation of insoluble minerals/compounds, and accumulation of soluble minerals (Mack et al. 1993).More wetland facies models are needed based on sedimentsaccumulating and soils forming in a range of landscape positions.

CASE STUDY: LOBOI SWAMP

Introduction and Methods

Loboi Swamp is a 1.5-km2 spring-fed wetland near the equator in theKenyan Rift Valley (Ashley et al. 2004). The climate is semiarid with

very high evapotranspirative losses (2500 mm/yr) that greatly exceedmean annual precipitation (750 mm/yr). The present wetlanddeveloped on a low-relief alluvial plain during the late Holocene (past700 years based on radiocarbon dating of seeds) (Fig. 9). Dominantvegetation of the wetland includes: Typha domingenesis Pers. (cattail, aC3 plant) and floating Cyperus papyrus L (papyrus, a C4 plant).Lowering of regional groundwater due to land use is currentlyshrinking the wetland substantially (Ashley et al. 2004).

A soil pit at the edge of Loboi Swamp, Kenya, was described byDriese et al. (2004) (Fig. 7A). Cores were obtained from the sameswamp in 2002 as part of a paleoclimate study of Pleistocene–Holocene paleowetlands, modern wetlands, and paleosols. Cores 1 and2 (;4 m apart) were collected within 2.5-m-tall Typha, at an elevationof 1023 m and with global positioning system (GPS) coordinates of00821.6260N, 036802.7650E, using a modified Livingston-type pistoncorer. Core 1 was used to study the sedimentology (Ashley et al. 2004),and core 2 was used to study paleosol development (reported here).The cores were split, described, scanned for magnetic susceptibility,and subsampled for thin sections, X-ray diffraction (XRD), and totalorganic carbon (TOC) analyses. One half of each core split waspreserved for archiving. Color was described using Munsellt soil colorcharts (Munsell 2000), and pedogenic features and classificationfollow USDA Soil Taxonomy (Soil Survey Staff 1998). Redoximor-phic features, pedal and other soil structures, and rooting weredescribed, after which each horizon was sampled for commercial thin-section preparation in order to: (1) identify characteristic wetland soilfeatures, (2) interpret the variability and relative timing of wetland soilfeatures vs. well-drained features (with depth), to determine thecumulative record of 700 years of wetland pedogenesis, and (3)identify pedogenic carbonates suitable for sampling for stable isotopeanalysis.

XRD at Baylor University was used to evaluate the mineralogy ofnodules present in the core, especially for distinguishing among calcite,ankerite, and siderite. Nodules are concentrations of soil constituents(Fe-Mn, CaCo3) that have no internal structure but have a sharplydefined boundary with surrounding soil, whereas concretions are a

FIG. 8.—Sedimentary facies model of a groundwater-fed wetland based on modern studies in northern Tanzania (modified from Deocampo

2002a). The three-dimensional block diagram depicts the various components of the groundwater-fed wetland model: spring, wetland, and

drainage system (channels and pools). This water source is permanent and thus home to hippopotami that require dependable water. Hippo

trails are a distinctive sedimentary structure associated with wetlands in Africa (Ashley and Liutkus 2002, Deocampo 2002b).



concentration of constituents that have distinctly banded, commonlyconcentric internal structure and a sharply defined boundary with thesurrounding matrix (Bullock et al. 1985). Glaebule is a generalizedterm for a concentration of soil constituents, generally Fe-Mn, having adiffuse boundary with the surrounding soil matrix (Brewer 1976).

Nodules were powdered using an opal mortar and pestle. Powderswere packed on a zero-background mount and scanned with a SiemensD5000 h-2h X-ray diffractometer. Representative thin sections of eachsoil horizon were prepared, with location and labeling for each thinsection shown in Figure 9 under ‘‘pedological analysis.’’ Thin-sectionsubscript labels correspond, roughly, to their relative positions in thecore segments: lower (l), middle (m), top (t). Thin sections wereexamined at Baylor University with an Olympus BX51 microscopeequipped with a 12.5 MPx digital camera. Micromorphologicaldescription of matrix terminology follows Brewer (1976), anddescriptions of all other features, including nodules and coatings,follow Stoops (2003). Abundance (few, common, many) and pedalstructure (wedge, subangular blocky, etc.) descriptions follow SoilSurvey Staff (1998).

Results

Soil Description: The soils sampled in Loboi Swamp Core #2classify as either Mollic Epiaquepts or Typic Epiaquepts in USDA SoilTaxonomy (Soil Survey Staff 1998), in contrast with the surfaceSulfaquepts and buried Aeric to Typic Tropaquepts described in theLoboi Swamp soil pit by Driese et al. (2004). Soil descriptions byhorizon for Loboi Swamp Core #2 are presented in Table 1.Descriptions of soil structure (peds) are difficult with a small-diametercore and would be improved with trenching. Overall, the soils areweakly to moderately developed; both the surface soil and buried soilclassify as Inceptisols, rather than Entisols, because of moderatelydeveloped horizonation and lack of diagnostic subsurface horizons.The A horizon was subdivided into A1 and A2 (Fig. 9) based upondifferences in root density. These A horizons, given more time for

organic matter accumulation, would likely have formed a Histicepipedon. The A2 horizon hosted very few soft carbonate masses;however, the soft masses increased towards the A2/Bw boundary. XRDanalyses were conducted on powdered hard carbonate nodules sampledfrom the Bw horizon and Bwb horizons. Both samples weredetermined to be nearly 100% calcite, with no XRD evidence forsiderite or significant Fe substitution for Ca as would occur in ankeriteor siderite.

Micromorphology: Micromorphological observations and all sig-nificant pedogenic attributes observed are summarized in Table 2. Allhorizons contained root traces with yellowish illuviated clay pore-linings (coatings), and opaque to reddish-brown redoximorphicfeatures such as Fe-Mn concentrations (nodules) and Fe-Mn coatingson grains and macropores. Pedogenic calcite was also ubiquitous inmost soil horizons, although total amounts, as well as hard nodules vs.soft masses, increased with depth. A pedogenic origin is the mostreasonable interpretation because the carbonate did not containlithologic or inherited carbonate constituents and showed evidence ofdesiccation during the time that it was formed. Sand and silt grainsconsisted of weathered volcanic rock fragments (VRFs), maficminerals such as pyroxenes and basaltic hornblende, as well asmonocrystalline quartz. The A1 horizon, defined in the core based onthe highest abundance of rooting, also had the highest abundance ofrooting in thin section, whereas the A2 horizon had a lower abundanceof rooting in thin section. The texture changed to slightly more siltyfrom the two A horizons to the B horizons, although the texture stillremained a silty clay. The Bwb horizon had all the pedogenic attributesof the Bw horizon; however, they were slightly less developed, andthere was notable preservation of some intervals of sediment withprimary bedding, some of which consisted of diatom layers.

Rooting was mainly identified by distinct yellowish clay-coatings onroot pores and by carbonate rhizoliths formed around roots (Fig. 10A,C, D). Microbanding of color and inclusions of Fe-Mn definedmultiple generations of clay illuviation (Fig. 10C, D). Detrital sand

FIG. 9.—Comparison of pedological and sedimentological analyses conducted on two juxtaposed cores (1 and 2) from Loboi Swamp, Kenya

(modified from Ashley et al. 2004). Soil and sediment cores are approximately correlative based upon thickness, texture, and facies. The

stratigraphy in core 1 reveals Loboi floodplain sediments overlain by incipient wetland deposits (organic-rich silt) capped with wetland peat.

Nitrogen (%) and LOI (loss on ignition) reflect increase in plant productivity toward the top, whereas decrease in magnetic susceptibility

reflects a dramatic reduction in minerals.



grains were common, ranging from weathered, subangular volcanicrock fragments (VRFs) to monocrystalline quartz, with a fewpedorelicts (reworked soil aggregates) (Fig. 10B). Calcium carbonatenodules (CaCO3) consisted of micrite recrystallized to microspar, withsparry calcite filling circumgranular cracks and septarian voids (Fig.10A). Carbonate nodules were commonly impregnated with Fe-Mnoxides and oxyhydroxides (Fig. 10E), and some surfaces of carbonatenodules were recrystallized to prismatic calcite. Fe-Mn diffuse mottlesand hard nodules were moderately to strongly developed, and Fecommonly coated detrital grains (Fig. 10B). Although both Fe and claycoatings were present, clay coatings were overall more dominant.Abrupt contacts in the basal part of the core (Bwb, lower) definedlayers rich in diatoms that were relatively unmodified by pedogenesis(Fig. 10F); however, diatoms were present in both the Bw and Bwbhorizons.

Interpretation

Overall, the soils are weakly to moderately developed, likely due to afluctuating water table and periodic saturated (aquic) conditionsdeveloped during the wet seasons. Micromorphologic investigationssupport the horizon designations proposed based upon macroscopicsoil core description (Fig. 9; Tables 1, 2). Both the A1 and A2 horizonshave the greatest abundance of root traces, as well as live roots. Claycoatings of root pores facilitated identification of root pores. The Bwand Bwb horizons showed an increase in pedogenic modificationmanifested by an increase in both Fe-Mn coatings and nodules, as wellas the development or ‘‘maturity’’ of pedological features, especiallythe laminated and/or crescentic clay coatings. The Bwb horizon, aburied soil horizon, has many similarities to the overlying A and Bwhorizons; however, as previously noted, root traces of live roots wereless abundant. Fe nodules and Fe impregnation of CaCO3 nodules inthe Bwb horizon, which resemble those present in the A and Bwhorizons, suggest similar redoximorphic conditions during theirdevelopment. The preservation of primary bedding in the Bwb horizonmay be a product of ephemeral flood events that introduced sedimentsand promoted diatom blooms at the onset of wetland formation circa700 year BP (cf. Ashley et al. 2004, Driese et al. 2004). The presence ofboth clay coatings and Fe-Mn coatings in all horizons suggestsepisodically fluctuating soil drainage conditions: Well-drained/dryconditions promote the development of clay coatings, whereas poorlydrained/wet conditions favor Fe-Mn coatings. Clay coatings had bothlaminated and crescent morphologies in the Bw and Bwb horizons,suggesting better or sustained well-drained conditions. Nodules of bothFe oxides (and Fe oxyhydroxides) and CaCO3 are additional evidenceof fluctuating drainage conditions. Both aquic conditions and theclayey texture of the soil generally do not promote the accumulation ofcalcium carbonate, and it appears likely that the carbonate has beeninfluenced by both pedogenic as well as phreatic (groundwater)processes; however, this remains for further stable isotope, cathodeluminescence, trace element, and micromorphology investigations.Interestingly, a slight change in texture does occur in the Bw and Bwbhorizons (mainly an increase in silt content), perhaps allowing forbetter drainage to form calcium carbonate nodules. The Bw horizonhas a greater silt content than that of the Bwb horizon, and has morecommon hard carbonate rhizoliths and nodules (Table 2).

Although ped structure was not easily observed in the B horizons,the juxtaposition of clay coatings, carbonates, and Fe-Mn coatings andmasses within all horizons supports a cumulative 700-year soil recordof multiple episodes of alternating saturation and aeration (cf. Driese etal. 2004). Loboi Swamp today experiences both a wet and dry season.However, the mean annual precipitation of 700 mm/yr is greatlyexceeded by evapotranspirative losses, which are about 2500 mm/yr(Ashley et al. 2004). Thus, it would be expected that the dry seasonwould likely leave the greatest ‘‘imprint’’ in soil morphology. In all

TA

BL

E1

.—D

escr

ipti

on

of

mac

rosc

op

icfe

atu

res,

Lo

bo

iS

wam

pC

ore

#2

.C

lass

ifica

tio

n:

Mo

llic

Ep

iaq

uep

to

rTy

pic

Ep

iaq

uep

t.

Ho

rizo

n

Dep

th

(cm

)C

olo

rT

extu

reC

on

cret

ion

sE

ffer

ves

cen

ceR

oo

tsB

ou

nd

ary

A1

0–

7D

ark

oli

ve

gra

y

(5Y

3/2

)

Cla

yey

,m

ois

t,ver

yfi

ne,

cru

mb

ly,

loo

se

No

ne

Man

yfi

ne

bla

cken

ed

frag

men

ts

Cle

ar,

stra

igh

t

A2

7–

48

Dar

ko

live

gra

y

(5Y

3/2

)

Cla

yey

,m

ois

t,ver

yfi

ne,

cru

mb

ly,

loo

se,

few

dis

tin

ct(7

.5Y

R5

/6)

soft

mas

ses

Few

,co

mm

on

No

ne

Fin

e,b

ifu

rcat

ing

Cle

ar,

stra

igh

t

Bw

48

–8

8V

ery

dar

kg

ray

ish

-

bro

wn

(2.5

Y3

/2)

Sil

tycl

ay,

mo

ist,

ver

yfi

ne,

sub

ang

ula

rblo

cky

;lo

ose

,

com

mo

n,

man

y,fi

ne

dis

tin

ct,

soft

mas

ses

(5Y

4/4

)

Co

mm

on

,fi

ne,

pro

min

ent,

calc

ium

carb

on

ate

con

cret

ion

s

(10

YR

7/6

)

Str

on

gF

ew,

fin

eG

rad

ual

,st

raig

ht

Bw

b8

8–

14

2D

ark

gre

en-g

ray

(2.5

Y3

/2)

Cla

yey

,m

ois

t,ver

yfi

ne,

sub

ang

ula

r

blo

cky,

loo

se

Few

,fi

ne,

pro

min

ent,

calc

ium

carb

on

ate

con

cret

ion

s(2

.5Y

7/6

)

Str

on

gF

ew,

fin

eN

ob

ou

nd

ary,

end

pro

file



TA

BL

E2

.—D

escr

ipti

on

of

mic

rosc

op

icfe

atu

res,

Lo

bo

iS

wam

pC

ore

#2

.

Ho

rizo

nC

olo

rP

eds

Ro

oti

ng

Co

atin

gs

No

du

les/

con

cret

ion

s

Pri

mar

yA

bu

nd

Mo

rph

olo

gy

Ab

un

dTy

pe

Mo

rph

olo

gy

Ab

un

dTy

pe

Mo

rph

olo

gy

/im

pre

g

A1

*o

bS

Am

Cla

y-c

oat

edp

ore

sf-

cC

lay

Co

atin

g-c

resc

ent/

men

isc

fC

aCO

3Ty

pic

w/O

Min

clu

sio

ns

N

fF

eC

oat

ing

-cap

pin

gc

Fe

Ty

pic

/str

on

g

mC

lay

Co

atin

gro

ot

po

res

fF

eCO

3Ty

pic

con

cen

tric

/ban

ded

A2

ob

SA

-SB

c-m

Cla

y-c

oat

edp

ore

sf-

cC

lay

Co

atin

g-c

resc

ent/

men

isc

fF

eA

mb

oid

al/s

tro

ng

N

mC

lay

Co

atin

g-n

on

lam

inat

edf

CaC

O3

Ty

pic

con

cen

tric

f-c

Fe

Hy

po

coat

ing

cF

eTy

pic

con

cen

tric

/str

on

g

(,2

00

lm)

A2

-Bw

lob

SA

-SB

f-w

Fe-

coat

edp

ore

sc

Cla

yC

oat

ing

/hy

po

coat

ing

sm

CaC

O3

Gla

ebu

les,

typ

icN

cC

lay

coat

edp

ore

sf

Fe

Hy

po

coat

ing

sc

Fe

Ty

pic

/st

ron

g-m

od

erat

e

fC

lay

Co

atin

g-c

resc

ent/

men

isc

fF

eTy

pic

/wea

k-m

od

erat

e

Bw

-tlb

ob

SA

-SB

c-m

Cla

y-c

oat

edp

ore

sm

Cla

yC

oat

ing

sc-

mC

aCO

3G

laeb

ule

s,ty

pic

N

fC

lay

Co

atin

gs-

lam

inat

edm

Fe

,1

00

lm

,st

ron

g(O

M?)

fC

lay

Hy

po

coat

ing

s

fF

eC

oat

ing

s

Bw

-mlb

ob

SA

cC

lay

-co

ated

po

res

cC

lay

Co

atin

gs

f-c

CaC

O3

Gla

ebu

les,

typ

icN

WE

fF

eH

yp

oco

atin

gs

mF

e,

10

0l

m,

stro

ng

(OM

?)

f-c

Fe

Inte

rcal

atio

ns

Bw

b-t

ob

SA

c-m

Cla

y-c

oat

edp

ore

sc-

mC

lay

Co

atin

gs

fC

aCO

3G

laeb

ule

s,ty

pic

N

WE

fC

lay

Co

atin

gs-

lam

inat

edc

Fe

Ty

pic

/mo

der

ate

Bw

b-m

ob

SA

cC

lay

-co

ated

po

res

f-c

Cla

yC

oat

ing

sf-

cC

aCO

3Ty

pic

con

cen

tric

N

fF

eTy

pic

/mo

der

ate

Bw

b-l

ob

?f

Cla

y-c

oat

edp

ore

sf

Cla

yC

oat

ing

sf

CaC

O3

Ty

pic

/am

bo

idal

Y?

fC

lay

Co

atin

gs-

lam

inat

ed

Ped

s:S

A¼

sub

ang

ula

rblo

cky,

SB¼

sub

rou

nd

ed,W

E¼

wed

ge.

Co

lor:

ob¼

oli

ve

bro

wn

,lo

b¼

lig

hto

live

bro

wn

,lb

ob¼

lig

htblu

eo

live

bro

wn

.A

bu

nd

ance

:f¼

fin

e,c¼

com

mo

n,m¼

man

y.*

All

ho

rizo

ns

are

silt

ycl

ayte

xtu

re.



FIG. 10.—Photomicrographs of Loboi Swamp Core LSA-2 (all photographs oriented with stratigraphic up toward top of page). A) Bwb horizon,

middle (m) sample showing micrite and microspar rhizolith formed around root, with internal porosity in center filled with calcite spar cement;

cross-polarized light (XPL). B) Bw horizon, top (t) sample, showing Fe-Mn concentrations (nodules) formed around weathered ferro-

magnesian silicate grains; plane-polarized light (PPL). C, D) PPL and XPL photo pair of Bwb horizon, lower (l) sample, showing microbanded

illuviated clay coating concentrated along floor of fine root pore—note alternating darker micro- to meso-scale bands of Fe-Mn indicating

episodes of alternating saturation and drainage. E) Bwb (l) horizon, Fe-Mn impregnation of interior of pedogenic micrite nodule following

circumgranular and septarian shrinkage cracks (PPL). F) Bwb (l) horizon, showing diatom-rich layer with preserved primary stratification,

which records flooding of wetland and standing water (PPL).



horizons, clay coatings and CaCO3 were more abundant thanpedological features indicative of poorly drained conditions.

Summary

(1) A1 and A2 horizons are characterized by abundant root traces,which are most easily identified by pedogenic clay coatings. (2)Pedogenic and groundwater(?) carbonates are exclusively calcite, withno evidence for siderite or ankerite. (3) Fe-Mn-oxide and oxyhydroxidecoatings and nodules were much more abundant in the Bw and Bwbhorizons, and (4) attributes of both well-drained (e.g., clay coatings,CaCO3 nodules) and poorly drained conditions (e.g., Fe-Mn coatings,Fe-Mn nodules) are present. However, indicators of well-drained soilconditions dominate. (5) The Bwb horizon and A1–A2 horizons havesimilar pedological attributes; however, the Bwb horizon also hasevidence of primary bedding. This structure is attributed to episodicflood events into the wetland and incomplete pedogenesis. (6) Thejuxtaposition of morphological and micromorphological featuresindicative of episodic drainage and aeration is attributed to waterinflow–outflow related to the unique hydrology of this Africangroundwater wetland.

DISCUSSION

Groundwater-fed wetlands need to be saturated for a significantportion of the year. In order to support a wetland, the hydrologic budgetcannot be negative for long because the water table would drop and thewetlands would dry up. If the budget were consistently positive, then astanding body of water like a pond or lake would grow until inflow wasbalanced by a spillway or groundwater recharge. Thus, groundwater-fed wetlands can only persist where multiple parameters areconsistently well balanced; the resulting record is intimately relatedto groundwater history. However, groundwater-fed wetlands do exist inmodern environments, and several have been interpreted from thegeologic record (Quade et al. 1995, Liutkus and Ashley 2003, Owen etal. 2009). They represent an end member of a continuum of palustrineenvironments that have varying numbers of days per year that the watertable is at the surface. Groundwater-fed wetlands, as opposed to riverfloodplains, are more likely to remain permanently saturated. They aremore likely to have fewer days or weeks per year with a depressedwater table than other scenarios because water supply in the subsurfaceis protected from evaporation and evapotranspiration and bufferedagainst short-term changes in water budget.

The waterlogged environment of a groundwater-fed wetlandproduces a distinctive record dictated by the aqueous geochemicalenvironment and the biota present. Both of these are controlled by theposition of the water table. Quade and colleagues were one of the firstresearch groups to interpret the geologic record produced on, and at thebase of, a hillslope by sustained groundwater seepage and existence of‘‘fossil springs’’ (Quade 1986, Quade and Pratt 1989 ). Groundwaterthat is undersaturated in carbonate will not leave the calcite ‘‘hallmark’’(i.e., tufa or travertine). Even when it is present, the carbonate may onlyindicate local areas of CO2 degassing proximal to discharge. Thus,clues need to be gathered from the mineral assemblages (Fig. 6), color,and extent of bioturbation. For example, Quade et al. (1995) reportedpale green (gleyed) mud containing hygrophilic fauna (terrestrialmolluscs needing damp ground) and a dense stand of phreatophytes.

Research on wetlands has tended to follow one of two parallel tracks:sedimentology or pedology. Sedimentologists tend to view the systemfrom bottom to top (old to young, a history of sedimentation), andpedologists tend to view the system from top to bottom (pedogenesisproceeds downward from the surface). The data collected and analysescarried out by the two approaches are both designed to determine thehistory of the wetland. The case study from Loboi Swamp, Kenya,provides an excellent example of the two different approaches applied

to the same record (Figs. 9, 10). The sedimentological approach yieldsa smooth record of wetland accretion. Whereas the pedologicalapproach provides more details of wetland processes, particularly therecord of water-table fluctuations. A combination of the sedimento-logical and pedological approaches leads to a better understanding ofthe record than either one could do on its own.

Inland wetlands, specifically groundwater-fed wetlands, are part of alarger landscape and will have adjacent sedimentary and soil records.Research should be carried out on them by adopting the facies analysesapproach that is so successful in sedimentology. Identifying buriedpaleosols and tracing them across topography to establish apaleocatena have not been easy; there are only few published examples(Deocampo et al. 2002, Kraus and Hasiotis 2006, Ashley et al. 2010a,Nordt et al. 2011). It should be possible if the appropriate parametersare chosen.

CONCLUSIONS

Groundwater-fed wetlands are ecosystems that range betweenephemerally wet to fully aquatic habitats; thus, the character of awetland record is directly related to the position of the water table overseasonal and longer timescales. Geochemical processes and biologicalprocesses are directly related to the position of the water table, and thishistory is recorded in horizon development, geochemical profiles, roottraces, color, and pedogenic structures. Wetland research has tended tofollow one of two tracks: sedimentology and pedology. The twoapproaches bring different perspectives to the interpretation of theirrecords. Sedimentologists tend to view the system from bottom to top(old to young, a history of sedimentation), and pedologists tend to viewthe system from top to bottom (pedogenesis proceeds downward fromthe surface). The wetland case study example presented here suggeststhat the pedological approach, because of the use of micromorphology,provides more details of the history of wetland processes, in particularthe record of water-table fluctuations. Water-table fluctuations areindicators of local water budget and in turn local climate. Thesedimentological approach yields a smooth record of wetland accretionreflecting the addition of mineral and organic matter over time. Acombination of the sedimentological and pedological approaches couldlead to a better understanding of the record than either one could do onits own. Finally, wetland paleosols should be studied in a larger spatialframework (paleocatenas) and longer temporal framework similar tothe facies models developed for sedimentary sequences.

ACKNOWLEDGMENTS

Research on core #1 was supported by National Science Foundation(NSF) grant EAR-0207705 to G.M.A. and V.C. Hover, RutgersUniversity. Research on the Loboi Swamp core #2 conducted at BaylorUniversity was supported by NSF grant EAR-0643152 awarded toS.G.D., and supported J.K.-R. We have profited by fruitful discussionswith Michelle Goman, Victoria Hover, Robin Renaut, Bernie Owen,Joseph Maitima Mworia, A.M. Muasya, and Steven Mathai. We aregrateful to Veronica Muiruri, National Museum of Kenya, forassistance with the palynology. We appreciate the insightful reviewsby Dennis Terry and Gary Weissmann and Volume Editor PaulMcCarthy.

REFERENCES

Armenteros I, Angeles Bustillo MA, Blanco JA. 1995. Pedogenic and

groundwater processes in a closed Miocene basin (northern Spain).

Sedimentary Geology 99:17–36.

Ashley GM, Barboni D, Domınguez-Rodrigo M, Bunn HT, Mabulla AZP, Diez-

Martin F, Barba R, Baquedano E. 2010a. Paleoenvironmental and

paleoecological reconstruction of a freshwater oasis in savannah grassland

at FLK North, Olduvai Gorge, Tanzania. Quaternary Research 74:333–343.



Ashley GM, Barboni D, Domınguez-Rodrigo M, Bunn HT, Mabulla AZP, Diez-

Martin F, Barba R, Baquedano E. 2010b. A spring and wooded habitat at FLK

Zinj and their relevance to origins of human behavior. Quaternary Research

74:304–314.

Ashley GM, Liutkus CM. 2002. Tracks, trails and trampling by large vertebrates

in a rift valley paleo-wetland, lowermost Bed II, Olduvai Gorge, Tanzania.

Ichnos 9:23–32.

Ashley GM, Maitima Mworia JM, Muasya AM, Owen RB, Driese SG, Hover

VC, Renaut RW, Goman MF, Mathai S, Blatt SH. 2004. Sedimentation and

evolution of a freshwater wetland in a semi-arid environment, Loboi Swamp,

Kenya, East Africa. Sedimentology 51:1–21.

Baas-Becking LGM, Kaplan IR, Moore D. 1960. Limits of the natural

environment in terms of pH and oxidation-reduction potentials. Journal of

Geology 68:243–284.

Beverly EJ. 2011. High-resolution paleoenvironmental and paleoclimatic

reconstruction of a Pleistocene catena and climosequence using paleopedol-

ogy and geochemistry of lake margin paleo-Vertisols, Olduvai Gorge,

Tanzania [unpublished MS thesis]: Rutgers University, New Brunswick, New

Jersey. 116p.

Blatt H, Middleton GV, Murray R. 1960. Origin of Sedimentary Rocks: Prentice-

Hall, Englewood Cliffs, New Jersey. 782 p.

Bradbury IK, Grace J. 1983. Primary production in wetlands. In Gore AJP

(Editor). Mires: Swamp, Bog, Fen, and Moor: Ecosystems of the World:

Elsevier, Amsterdam. p. 285–310.

Brantley SL, Megonigal JP, Scatena FN, Balogh-Brunstad Z, Barnes RT, Bruns

MA, Van Cappellen P, Dontsova K, Hartnett HE, Hartshorn AS, Heimsath A,

Hendon E, Jin L, Keller CK, Leake JR, McDowell WH, Meinzer FC,

Mozdzer TJ, Petsch S, Pett-Ridge J, Pregitzer KS, Raymond PA, Riebe CS,

Shumaker K, Sutton-Grier A, Walter R, Yoo K. 2011. Twelve testable

hypotheses on the geobiology of weathering. Geobiology 9:140–165.

Brewer R. 1976. Fabric and Mineral Analysis of Soils: Krieger Publishers, New

York. 482 p.

Bullock P, Fedoroff N, Jongerius A, Tursina T, Babel U. 1985. Handbook of Soil

Thin Section Description: Waine Research, Albrighton, UK. 152 p.

Charman DJ. 2001. Biostratigraphic and palaeoenvironmental applications of

testate amoebae. Quaternary Science Reviews 20:1753–1764.

Coleman JM. 1988. Dynamic changes and processes in the Mississippi River

Delta. Geological Society of America Bulletin 100:999-1015.

Cowardin LM, Carter V, Golet FC, LaRoe EY. 1979. Classification of Wetlands

and Deepwater Habitats of the United States: US Fish and Wildlife Service,

Office of Biological Service, Washington, DC. p. 103.

Das BK. 2002. Biogeochemistry as an indicator of organic matter sources,

paleolimnological and paleoenvironmental changes in lacustrine sediments—

a study of two Himalayan lakes. Environmental Geosciences 9:115–126.

Deocampo DM. 2002a. Sedimentary processes and lithofacies in lake-margin

groundwater-fed wetlands in East Africa. In Renaut RW, Ashley GM

(Editors). Sedimentation in Continental Rifts, Special Publication 73: SEPM

Society for Sedimentary Geology, Tulsa, Oklahoma. p. 295–308.

Deocampo DM. 2002b. Sedimentary structures generated by Hippopotamus

amphibius in a lake-margin wetland, Ngorongoro Crater, Tanzania. Palaios

17:212–217.

Deocampo DM. 2004. Hydrogeochemistry in the Ngorongoro Crater, Tanzania,

and implications for land use in a World Heritage Site. Applied Geochemistry

19:755–767.

Deocampo DM. 2010. The geochemistry of continental carbonates. In Alonso-

Zarza AM, Tanner LH (Editors). Carbonates in Continental Settings:

Geochemistry, Diagensis and Applications, Developments in Sedimentology

62: Elsevier, Amsterdam. p. 1–59.

Deocampo DM, Ashley GM. 1999. Siliceous islands in a carbonate sea: modern

and Pleistocene spring-fed wetlands in Ngorongoro Crater and Oldupai

Gorge, Tanzania. Journal of Sedimentary Research 69:974–979.

Deocampo DM, Blumenschine RJ, Ashley GM. 2002. Wetland diagenesis and

traces of early hominids, Olduvai Gorge. Quaternary Research 57:271–281.

Driese SG, Ashley GM, Li Z-H, Hover VC, Owen RB. 2004. Possible Late

Holocene equatorial palaeoclimate record based upon soils spanning the

Medieval Warm Period and Little Ice Age, Loboi Plain, Kenya. Palae-

ogeography, Palaeoclimatology, Palaeoecology 213:231–250.

Driese SG, Ludvigson GA, Roberts JA, Fowle DA, Gonzalez LA, Smith JJ,

Vulava VM, McKay LD. 2010. Micromorphology and stable-isotope

geochemistry of historical pedogenic siderite formed in PAH-contaminated

alluvial clay soils, Tennessee, USA. Journal of Sedimentary Research

80:943–954.

Faure G. 1991. Principles and Applications of Inorganic Chemistry: MacMillan

Publishing Company, New York. 626 p.

Garrels RM, Christ CL. 1965. Solutions, Minerals, and Equilibria: Harper and

Row, New York. 450 p.

Garrels RM, Mackenzie FT. 1967. Origin of the Chemical Composition of

Springs and Lakes; Equilibrium Concepts in Natural Water Systems,

Advances in Chemistry 67: American Chemical Society, Washington, DC.

p. 222–242.

Gell P. 2010. With the benefit of hindsight: the utility of paleoecology in wetland

condition assessment and identification of restoration targets. In Batty LC,

Hallberg KB (Editors). Ecology of Industrial Pollution; Ecological Reviews:

Cambridge University Press, Cambridge, UK. p. 162–188.

Grosjean M. 2001. Mid-Holocene climate in the south-central Andes: Humid or

dry? Science 292:2391.

Hasiotis ST. 1997. Redefinition of continental ichnology and the Scoyenia

Ichnofacies [unpublished PhD thesis]: University of Colorado, Boulder. 182 p.

Hasiotis ST. 2000. The invertebrate invasion and evolution of Mesozoic soil

ecosystems: the ichnofossil record of ecological innovations. In Gastaldo RA,

DiMichele WA (Editors). Phanerozoic Terrestrial Ecosystems, Short Course

6: The Paleontological Society, New Haven, Connecticut. p. 141–169.

Hasiotis ST, Kraus MJ, Demko TM. 2007. Climate controls on continental trace

fossils. In Miller WI (Editor). Trace Fossils: Concepts, Problems, and

Prospects: Elsevier, Amsterdam. p. 172–195.

He X, Vepraskas MJ, Lindbo DL, Skaggs RW. 2003. A method to predict soil

saturation frequency and duration from soil color. Soil Science Society of

America Journal 67:961–969.

Ho C, Coleman JM. 1969. Consolidation and cementation of recent sediments

in the Atchafalaya Basin. Geological Society of America Bulletin 80:183–

192.

Holliday VT. 2004. Soils in Archaeological Research: Oxford University Press,

New York. 464 p.

Hover VC, Ashley GM. 2003. Geochemical signatures of paleodepositional and

diagenetic environments: a STEM/AEM study of authigenic clay minerals

from an arid rift basin, Olduvai Gorge, Tanzania. Clays and Clay Minerals

51:231–251.

Hurt GW, Whited PM, Pringle RF. 1998. Field Indicators of Hydric Soil in the

United States, Version 4.0: US Department of Agriculture, Natural Resources

Conservation Service, Ft. Worth, Texas. 30 p.

Johnson CR, Ashley GM, De Wet CB, Dvoretsky R, Park L, Hover VC, Owen

RB, McBrearty S. 2009. Tufa as a record of perennial fresh water in a semi-

arid rift basin, Kapthurin Formation, Kenya. Sedimentology 56:1115–1137.

Jones BF, Deocampo DM. 2003. Geochemistry of saline lakes. In Drever JI

(Editor). Surface and Groundwater, Weathering, and Soils, Treatise on

Geochemistry 5: Elsevier, Amsterdam. p. 393–424.

Keddy PA. 2000. Wetland Ecology: Principles and Conservation: Cambridge

University Press, Cambridge, UK. 632 p.

Klappa CF. 1980. Rhizoliths in terrestrial carbonates: classification, recognition,

genesis and significance. Sedimentology 27:613–629.

Kraus MJ. 1999. Paleosols in clastic sedimentary rocks: their geologic

applications. Earth-Science Reviews 47:41–70.

Kraus MJ, Hasiotis ST. 2006. Significance of different modes of rhizolith

preservation to interpreting paleoenvironmental and paleohydrologic settings:

examples from Paleogene paleosols, Bighorn Basin, Wyoming, USA.

Journal of Sedimentary Research 76:633–646.

Krumbein WC, Garrels RM. 1952. Origin and classification of chemical

sediments in terms of pH and oxidation-reduction potentials. Journal of

Geology 60:1–33.

Lin H, Bouma J, Pachepsky Y, Western A, Thompson J, van Genuchten R, Vogel

H-J, Lilly A. 2006. Hydropedology: synergistic integration of pedology and

hydrology. Water Resources Research 42:W05301, doi:10.1029/

2005WR004085

Liutkus CM, Ashley GM. 2003. Facies model of a semiarid freshwater wetland,

Olduvai Gorge, Tanzania. Journal of Sedimentary Research 73:691–705.

Liutkus CM, Wright JD, Ashley GM, Sikes NE. 2005. Paleoenvironmental



interpretation of lake-margin deposits using d13C and d18O results from Early

Pleistocene carbonate rhizoliths, Olduvai Gorge, Tanzania. Geology 33:377–

380.

Mack GH, James WC, Monger HC. 1993. Classification of paleosols.

Geological Society of America Bulletin 105:129–136.

Marchant R, Taylor D. 1998. Dynamics of montane forest in central Africa

during the late Holocene: a pollen-based reecord from western Uganda. The

Holocene 8:375–381.

Marshak S. 2004. Essentials of Geology: W.W. Norton & Company, New York.

628 p.

Melchor RN, Genise JF, Miquel SE. 2002. Ichnology, sedimentology and

paleontology of Eocene calcareous paleosols from a palustrine sequence,

Argentina. Palaios 17:16–35.

Mitsch WJ, Gosselink JG. 2000. Wetlands: John Wiley & Sons, New York.

920 p.

Munsell. 2000. Munsell Soil Color Charts: GretagMacbeth, New Windsor, New

York.

Nordt LC, Dworkin SI, Atchley SC. 2011. Ecosystem response to soil

biogeochemical behavior during the Late Cretaceous and early Paleocene

within the western interior of North America. Geological Society of America

Bulletin 123:1745–1762.

Norstrom E, Scott L, Partridge TC, Risberg J, Holmgren K. 2009.

Reconstruction of environmental and climate changes at Braamhoek wetland,

eastern escarpment South Africa, during the last 16,000 years with emphasis

on the Pleistocene–Holocene transition. Palaeogeography, Palaeoclimatol-

ogy, Palaeoecology 271:240–258.

NRC. 2001. Basic Research Opportunities in Earth Science: National Research

Council, Washington, DC. 154 p.

Olago DO, Street-Perrott FA, Perrott RA, Odada EO. 2003. Late Holocene

sedimentology and palaeoenvironment of Kiluli Swamp, Mount Kenya.

African Journal of Science and Technology (AJST) 4:12–33.

Owen RB, Renaut RW, Hover VC, Ashley GM, Muasya AM. 2004. Swamps,

springs, and diatoms: wetlands of the semi-arid Bogoria–Baringo Rift, Kenya.

Hydrobiologia 518:59–78.

Owen RB, Renaut RW, Scott JJ, Potts R, Behrensmeyer AK. 2009. Wetland

sedimentation and associated diatoms in the Pleistocene Olorgesailie Basin,

southern Kenya Rift Valley. Sedimentary Geology 222:124–137.

Prothero DR, Schwab F. 2004. Sedimentary Geology: An Introduction to

Sedimentary Rocks and Stratigraphy: W.H. Freeman, New York. 575 p.

Quade J. 1986. Late Quaternary environmental changes in the Upper Las Vegas

Valley, Nevada. Quaternary Research 26:340–357.

Quade J, Mifflin MD, Pratt WL, McCoy W, Burckle L. 1995. Fossil spring

deposits in the southern Great Basin and their implications for changes in

water-table levels near Yucca Mountain, Nevada, during Quaternary time.

Geological Society of America Bulletin 107:213–230.

Quade J, Pratt WL. 1989. Late Wisconsin groundwater discharge environments

of the southwestern Indian Springs Valley, southern Nevada. Quaternary

Research 31:351–370.

Reading HG. 1996. Sedimentary Environments and Facies: Blackwell, Oxford,

UK. 688 p.

Reeves JM, De Deckker PD, Halse SA. 2007. Groundwater ostracods from the

arid Pilbara region of northwestern Australia: distribution and water

chemistry. Hydrobiologia 585:99–118.

Reineck HE, Singh IB. 1975. Depositional Sedimentary Environments:

Springer-Verlag, New York. 551 p.

Renaut R. 1993. Zeolite diagenesis of late Quaternary fluviolacustrine

sediments and associated calcrete formation in the Lake Bogoria Basin,

Kenya Rift Valley. Sedimentology 40:271–301.

Retallack GJ. 1997. A Color Guide to Paleosols: Wiley, Chichester, UK. 175 p.

Retallack GJ. 2001. Soils of the Past: An Introduction to Paleopedology:

Blackwell Science, Oxford, UK. 404 p.

Rucina SM, Muiriri VM, Downton L, Marchant R. 2010. Late-Holocene

savanna dynamics in the Amboseli Basin, Kenya. The Holocene 20:667–677.

Schwalb A, Burns SJ, Cusminsky G, Kelts K, Markgraf V. 2002. Assemblage

diversity and isotopic signals of modern ostracodes and host waters from

Patagonia, Argentina. Palaeogeography, Palaeoclimatology, Palaeoecology

187:323–339.

Sheldon ND, Tabor NJ. 2009. Quantitative paleoenvironmental and paleocli-

matic reconstruction using paleosols. Earth-Science Reviews 95:1–52.

Shilling AM. 2012. Location and Longevity of Groundwater-Fed Wetlands,

Lake Eyasi Basin, Tanzania: Rutgers University, New Brunswick, New

Jersey. 126p.

Singer A. 1980. The palaeoclimate interpretation of clay minerals in soil and

weathering profiles. Earth-Science Reviews 15:303–326.

Soil Survey Staff. 1998. Keys to Soil Taxonomy: US Department of Agriculture,

Natural Resources Conservation Service, Washington DC. 330 p.

Stewart WN, Rothwell GW. 1993. Paleobotany and the Evolution of Plants:

Cambridge University Press, New York. 536 p.

Stoops G. 2003. Guidelines for Analysis and Description of Soil and Regolith

Thin Sections: Soil Science Society of America, Madison, Wisconsin. 184 p.

Stoops G, Eswaran H. 1985. Morphological characteristics of wet soils. In

Wetland Soils: Characterization, Classification, and Utilization: International

Rice Research Institute, Los Banos, Philippines. p. 177–189.

Thorp J. 1949. Effects of certain animals that live in soils. Scientific Monthly

68:189–191.

Tooth S, McCarthy TS. 2007. Wetlands in drylands: geomorphological and

sedimentological characteristics, with emphasis on examples from southern

Africa. Progress in Physical Geography 31:3–41.

Valentine KWG, Dalrymple JB. 1976. Quaternary buried paleosols: a critical

review. Quaternary Research 6:209–222.

van der Kamp G, Hayashi M. 2009. Groundwater-wetland ecosystem interaction

in the semiarid glaciated plains of North America. Hydrogeology Journal

17:203–214.

Vepraskas MJ. 1994. Redoximorphic Features for Identifying Aquic Conditions:

North Carolina Agricultural Research Service, Raleigh, North Carolina.

Technical Bulletin 301. 33 p.

Vepraskas MJ. 2001. Morphological features of seasonally reduced soils. In

Vepraskas MJ, Richardson JL (Editors). Wetland Soils: Genesis, Hydrology,

Landscapes, and Classification: Lewis Publishers, Boca Raton, Florida. p.

163–182.

Vepraskas MJ, Guertal WR. 1991. Morphological indicators of soil wetness. In

Kimble JM (Editor). Classification and Management of Wet Soils: US

Department of Agriculture, Soil Conservation Service, Washington, DC. p.

307–312.

Walker RG (Editor). 1984. Facies Models, Geoscience Canada Reprint 1:

Geological Association of Canada, Toronto, Ontario. 317 p.

Wallwork JA. 1970. Ecology of Soil Animals: McGraw-Hill, London. 283 p.

Wilding LP, Lin H. 2003. Advancing the frontiers of soil science towards a

geoscience. Geoderma 131:257–274.

Wust RAJ, Bustin RM. 2004. Late Pleistocene and Holocene development of

the interior peat-accumulating basin of tropical Tasek Bera, Peninsular

Malaysia. Palaeogeography, Palaeoclimatology, Palaeoecology 211:241–

270.

Ziegler H. 1995. Stable isotopes in plant physiology and ecology. In Behnke H-

D (Editor). Progress in Botany: Springer-Verlag, Berlin. p. 1–24.



groundwater-fed wetland sediments and...

Documents