Pedogenic processes and domain boundaries in a Vertisol climosequence: evidence from titanium and zirconium distribution and morphology

www.elsevier.com/locate/geoderma

Geoderma 116 (2003) 279–299

Pedogenic processes and domain boundaries in a

Vertisol climosequence: evidence from titanium and

zirconium distribution and morphology

Cynthia A. Stiles*, Claudia I. Mora, Steven G. Driese

Department of Geological Sciences, The University of Tennessee, Knoxville, TN 37996-1410, USA

Received 7 March 2002; accepted 17 February 2003

Abstract

The occurrences of titanium (Ti) and zirconium (Zr) within eight Vertisols formed in a

climosequence on the Upper Beaumont Formation of the Texas Gulf Coastal Plain were investigated

in order to determine processes responsible for Ti and Zr redistribution during pedogenesis.

Discontinuities defined by significant shifts in particle size distribution and the content (in volume

percent) of Zr are present at 160 to 260 cm depth in each pedon. The discontinuities are interpreted to

be functional boundaries, i.e., physico-chemical expressions of pedogenic domains, between an

upper soil domain dominated by open-system pedogenesis and a lower, more closed-system domain

dominated by chemical weathering. The depth at which the functional boundary occurs is dependent

on physical and hydrogeochemical influences, which are largely a function of available water. Soil

materials above the discontinuities are slightly coarser textured and enriched in Zr, whereas below

the sediments are finer textured and have lower and more constant Zr contents. The Zr is associated

almost exclusively with zircon and Zr contents correlate positively to the weight percent

sand + coarse silt, with negligible Zr present in the < 20 Am size fraction. By comparison, Ti

occurs preferentially in the < 20 Am size fraction and there are no marked discontinuities in Ti

contents with depth. Scanning electron microscopy (SEM) of individual zircon and rutile grains

show predominantly physical damage to zircons, whereas rutile grains appear to have been affected

predominantly by chemical weathering. Thus, different processes dictate Ti and Zr content and

distribution in these Vertisols, although both elements are often considered immobile in weathering

profiles.

Profile volume loss/gain (i.e., soil strain, e), a mass-balance calculation that assumes Zr or Ti

immobility during pedogenesis, indicate eZr values nearly four times greater than eTi. Large values of

0016-7061/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved.

doi:10.1016/S0016-7061(03)00105-8

* Corresponding author. Present address: Department of Soil Science, University of Wisconsin-Madison,

Madison, WI 53706-1299, USA. Tel.: +1-608-262-0331; fax: +1-608-265-2565.

E-mail addresses: [email protected] (C.A. Stiles), [email protected] (C.I. Mora), [email protected]

(S.G. Driese).

C.A. Stiles et al. / Geoderma 116 (2003) 279–299280

eZr within the upper soil domains are due primarily to sand and coarse silt additions to the Vertisols

and preclude use of Zr as a basis of mass-balance calculations in these soils, despite its relative

chemical stability. By comparison, Ti is generally conserved within the clay-rich soil profiles and is

therefore better suited for mass-balance calculations of volume change and mobile element

translocation during pedogenesis.

D 2003 Elsevier Science B.V. All rights reserved.

Keywords: Vertisols; Mass-balance; Strain; Pedogenic domains

1. Introduction

In weathering profiles, there is an important transition at depth between the open-

system pedogenesis that characterizes surface soils [i.e., those described in standard

USDA-Natural Resource Conservation Service (NRCS) soil surveys] and the relatively

closed system of the deeper soil and underlying regolith from which the soil continues to

form. This transition may, in some cases, be equivalent to a ‘‘lithologic discontinuity’’, a

term used by field pedologists to describe a distinct change in particle-size distribution,

grain morphology, lithologic composition, and/or color, and which is often expressed as

well in the bulk geochemistry of the profile. The term ‘‘functional boundary’’ is here used

to denote compositional differences in weathering profiles, whether they arise from

pedogenic or sedimentological factors. Unfortunately, there are no firm criteria to differ-

entiate between geochemical differences arising from an actual shift in the parent lithology

(i.e., mantled soils) from a weathering front or physical and chemical boundary between

domains of a single soil experiencing different pedogenic processes (Soil Survey Staff,

1998).

Brimhall et al. (1991) recognized two distinct layers in a lateritic bauxite (Jarrahdale,

Western Australia) weathering profile that had previously been described as a sedimento-

logical discontinuity. The upper (superactive) layer was dominated by bioturbation and

downward translocation of authigenic and exotic materials. The lower (subactive) zone

experienced some volume loss resulting from weathering and removal of various

components in solution, however its relatively low permeability limited invasive trans-

location. The functional boundary (‘‘translocation crossover’’; Brimhall et al., 1991)

between layers in these ‘‘old’’ (>1 Ma) soils occurs at around 3.5 m depth. In this example,

the functional boundary in the weathering profile reflected the effects of gross differences

in pedogenic processes acting on a single parent material rather than actual difference in

parent material composition. We introduce the equivalent term ‘‘pedogenic domains’’ for

the upper and lower zones described in Brimhall et al. (1991). Pedogenic domains are used

to describe portions of weathering profiles that experience differing intensities and

mechanisms of pedogenic processing.

Functional boundaries in soil profiles can be identified by the distribution of

‘‘immobile’’ elements in the profile. Immobile elements are residually enriched relative

to the more mobile elements released from soluble mineral phases in leached zones in the

profiles. Both Ti and Zr are commonly considered immobile in weathering environments

owing to the relatively insoluble nature of the detrital minerals in which they are

C.A. Stiles et al. / Geoderma 116 (2003) 279–299 281

concentrated, such as zircon (ZrSiO4) and rutile/anatase (TiO2; Marshall and Haseman,

1942; Jackson and Sherman, 1953). If they are both immobile, the ratio Ti:Zr should

remain constant throughout profile depth. Indeed, Ti:Zr ratios are a proxy for parent

material uniformity (Reheis, 1990; Birkeland, 1999).

Both Ti and Zr can be problematic as immobile index elements because the minerals

containing them may be, in some circumstances, transported or chemically reactive.

Zirconium is almost exclusively associated with detrital zircon, an extremely durable

mineral that can chemically survive numerous recycling events. Small additions of zircon

to the upper portions of a soil, commonly introduced by aeolian or volcanic processes,

can drastically change both the total Zr contents (wt.%) and Zr distribution within soil

profiles (Brimhall et al., 1991). Titanium can occur in a number of phases, both of

Fig. 1. Modern Vertisol climosequence sampling locations by series and pedon number.


detrital or pedogenic origin (Kaup and Carter, 1987; Fitzpatrick and Chittleborough,

2002). For example, Ti may replace iron (Fe) in hydrous Fe oxyhydroxide phases,

particularly in finer-textured soils, leading to the ‘‘leucoxene’’-weathering phase assem-

blage (most resistant phases or Stage 13; Jackson and Sherman, 1953). Despite the

chemical reactivity of some Ti-bearing detrital phases compared to zircon, Ti is still

generally conserved in the bulk soil matrix, particularly in fine-textured, Fe-enriched

matrices (Driese et al., 2000). Because of their sometimes contrasting behavior, i.e., the

sensitivity of Zr to open-system physical processes and of Ti to relatively conservative

hydrochemical enrichment, these elements may be used not only as a measure of parent

material uniformity, but also as indicators of soil domains dominated by physical or

chemical pedogenic processes.

Within the context of a climosequence of Vertisols from the Texas Gulf Coastal

Plain, the primary objectives of this study are to: (1) delineate Ti and Zr composi-

Table 1

Soil taxonomic, geographic, and climatic data for pedons used in this studya

Series name and U.S. Soil

Taxonomy descriptionbProfile

designationcGeographic

location

Ambient temperature

(jC)Precipitation

(mm)

MATd Highe Lowe MAP Annual deficitf

League

(fine, smectitic,

hyperthermic Oxyaquic

Hapluderts)

LEG 245A

(S00-TX-245-1)

30j02V22WN94j11V36WW

20.1 25.5 14.7 1437 604

Lake Charles

(fine, smectitic,

hyperthermic, Typic

Hapluderts)

LAC 201

(S99-TX-201-1)

LAC 157

(S99-TX-157-1)

LAC 481

(S99-TX-481-1)

29j35V40WN95j04V14WW29j24V12WN95j43V42WW29j22V21WN96j04V22WW

20.6

20.6

21.0

25.8

26.3

25.2

15.4

14.8

16.8

1321

1170

1124

774

866

1019

Laewest

(fine, smectitic,

hyperthermic, Typic

Hapluderts)

LAW 239

(S00-TX-239-1)

LAW 469

(S99-TX-469-1)

LAW 391

(S99-TX-391-1)

28j52V48WN96j24V11WW28j43V12WN96j45V23WW28j28V27WN97j07V00WW

21.0

21.2

21.8

25.2

26.7

27.4

16.8

15.8

16.1

1066

1000

924

1149

1381

1402

Victoria

(fine, smectitic,

hyperthermic Calcic

Haplusterts)

VIC 409

(S99-TX-481-3)

28j06V45WN97j20V55WW

21.8 27.4 16.1 844 1482

a All climate information derived from 50 to 100 years of daily temperature and precipitation data from

nearest Class 1 meteorological recording station to sampling locations; National Climate Data Center, Asheville,

NC.b Detailed information on these soil series is available at http://ortho.ftw.nrcs.usda.gov/osd/osd.html.c Detailed profile descriptions and characterization data are available at http://vmhost.cdp.state.ne.us:96/

METHD.HTML using the soil survey sample number given in parentheses.d Mean annual temperature.e Mean values.f Shown as 0� (MAP� potential evapotranspiration) such that deficits are positive values.

http:\\www.ortho.ftw.nrcs.usda.gov\osd\osd.html

http:\\www.vmhost.cdp.state.ne.us:96\METHD.HTML


tional differences with depth; (2) determine the depths to functional boundaries in the

pedons, and (3) assess if the compositional and textural differences between the

superactive (upper) and subactive (lower) units are due strictly to pedogenic processes

or represent actual lithologic differences within profiles. Morphological surveys and

statistical comparison of particle size and compositional trends are used to develop an

interpretive model of Vertisol pedogenesis and to augment considerations for choice of

an appropriate immobile index element for mass-balance calculations in fine-textured

soils.

2. Materials and methods

2.1. Geographic setting and sampling

The climosequence examined in this study developed on a coast-parallel terrace

exposing the upper units of the Pleistocene-age (35 ka) Beaumont Formation, which

Fig. 2. Ti/Zr ratios for climosequence microlow pedons.


consists of mixed fluvio-deltaic siliciclastic sediments extending from the Louisiana–

Texas border to Mexico (Bernard and LeBlanc, 1965; Blum and Price, 1994; Birdseye

and Aronow, 1991). These deposits are dissected by several rivers flowing into the Gulf

of Mexico. Sampling of younger terraces and reworked soils were avoided through

reconnaissance soil cores, and examination of aerial photos and soil maps. Eight pedons

representing four soil series were used in this study (League, Lake Charles, Laewest, and

Victoria); their MAP values range from 844 to 1437 mm (Fig. 1). Geographic,

taxonomic, and climate information for each location is listed in Table 1. Samples were

recovered from large soil pits (2 m wide, 3–5 m long) that transected microtopographic

features peculiar to Vertisols, also known as ‘‘gilgai’’ microtopography (Lynn and

Williams, 1992), herein referred to as microhighs and microlows. Bulk soil samples

from microtopographic pairs were collected at 10 cm intervals from the surface to the

base of each pit for geochemical analysis, and from each pedogenic horizon for bulk

density measurement.

Table 2

Climosequence functional boundary depths, Ti/Zr volumetric contents, percent differences, and statistical

information for the upper and lower units in microlow profiles

Element Property Pedon

LEG 245A LAC 201 LAC 157 LAC 481 LAW 239 LAW 469 LAW 391 VIC 409

FBa depth (cm) 200 190 170 160b 160 170 220 260

Ti UUc composition

(mg cm� 3)

10.47

(0.21)d9.04

(0.31)

9.48

(0.21)

9.55

(0.21)

9.37

(0.11)

7.87

(0.17)

7.22

(0.10)

6.13

(0.14)

LU composition

(mg cm� 3)

9.60

(0.78)

8.39

(0.44)

8.49

(0.37)

7.86

(0.21)

7.12

(0.27)

7.52

(0.39)

6.56

(0.07)

5.48

(0.26)

% Difference � 9.1 � 7.7 � 11.7 � 21.5 � 31.6 � 4.7 � 9.9 � 12.0

Mean

separation

0.88

(2.55)e0.65

(2.44)

0.99

(2.29)

1.69

(2.10)

2.25

(1.62)

0.35

(2.12)

0.65

(1.74)

0.65

(1.36)

Significance nsf ns ns ns * ns ns ns

Zr UU composition

(mg cm� 3)

1.45

(0.05)

1.46

(0.19)

0.99

(0.06)

0.79

(0.05)

1.38

(0.05)

1.63

(0.098)

0.83

(0.05)

1.59

(0.13)

LU composition

(mg cm� 3)

0.74

(0.14)

0.45

(0.09)

0.26

(0.06)

0.26

(0.01)

0.34

(0.14)

0.55

(0.10)

0.47

(0.02)

0.51

(0.32)

% Difference � 82.7 � 220.1 � 239.3 � 179.4 � 254.9 � 167.3 � 75.5 � 210.7

Mean

separation

0.66

(0.32)

1.03

(0.37)

0.70

(0.15)

0.51

(0.15)

1.00

(0.20)

1.03

(0.30)

0.36

(0.23)

1.08

(0.44)

Significance ** ** ** ** ** ** ** **

a Functional boundary.b Mean depth of irregular Ti/Zr composition.c UU= super-active domain, LU= sub-active domain.d Parenthetical value in compositional field is the 95% confidence interval for the individual profile data set.e Mean separation calculated as least significant difference at 95%= ta/2(2s

2/r)1/2, where ta/2 is Student’s

numerical probability coefficient, s2 is sample set variance, and r is the number of samples in each profile data set.

Parenthetical value is the mean separation value for that profile.f ns =Not significant.

*Significant at Pz 97.5%.

**Significant at Pz 99%.


2.2. Analytical methods

Whole soil samples were oven-dried at 60 jC, ground in a shatterbox, and pressed

powder pellets were used for bulk geochemical analysis using X-ray fluorescence (XRF;

EG&G ORTEC TEFA III, using methods described in Singer and Janitsky, 1986).

Precision for TiO2 and Zr is 0.01 wt.% and 23 mg kg� 1, respectively. Bulk density

was determined by the paraffin-coated air-dried clod method (Blake and Hartge, 1986).

Because significant carbonate content can lower the composite sample weight, elemental

weight percentages were corrected for carbonate content prior to mass-balance calculations

(modified from Soil Survey Staff, 1995). Direct volumetric comparisons are accomplished

by multiplying the wt.% by bulk density. Particle size was analyzed by the pipette method

(Gee and Bauder, 1986) at the National Soil Survey Laboratory in Lincoln, NE. Size

fraction separations of pedons LAC 481 and LAW 239 were made by wet sieving of bulk

samples shaken overnight in buffered Na-hexametaphosphate (10 g/l) solution to separate

the sand fraction (0.05–2 mm). Coarse (20–50 Am), medium to fine (5–20 Am), and very

fine (2–5 Am) silts were separated by centrifugation (Jackson, 1985) and chemically

analyzed by XRF, as described above.

Heavy mineral grains were separated from very fine sand fractions (125–62.5 Am)

above and below the functional boundaries in six microlow profiles using sodium

metatungstate. Zircon and rutile grains were isolated from other heavy minerals based

on their relative relief under the binocular scope, and from each other based on their

different refractive indices. Digital images of at least 25 grains from each sample were

made using a Hitachi S-3500 N scanning electron microscope and surveyed for various

surface features, as described by Darmody (1985) and Tejan-Kella et al. (1990). Two

Ti (mg cm-3)

0 5 10 15

Ti:

Zr

0

5

10

15

20

25

30

35

40

A.

Zr (mg cm-3)

0 1 2 3

Zr:

Ti

0.0

0.1

0.2

0.3

0.4 Above FBBelow FB

B.

Fig. 3. Microlow profile volumetric elemental content/ratio comparisons: (A) Ti; (B) Zr.


categories of surface features were utilized as evidence for process: (1) dissolution and

edge attrition, surface roughness and scaling, and oriented or random etch pits as

indicators of hydrogeochemical alteration; (2) percussion marks (conchoidal fracture/

breakage blocks/v-shaped pits), hairline cracks, and edge-rounding/frosting as indicators

of physical transport. Results were tabulated for the profiles as a mean percentage of grains

exhibiting these respective features for each category.

3. Results

3.1. Bulk geochemical analysis

We confine the majority of our discussion to observations from the microlows. General

pedogenic models of Vertisols indicate that seasonal shrink–swell induces physico-

chemical transport along slickenside planes, thereby creating and sustaining elevated

Silt Separate (wt %

)

0

10

20

30

40

50

60

70

80

90

100

Cla

y Se

para

te (w

t %)

0

10

20

30

40

50

60

70

80

90

100

Sand Separate (wt %)0102030405060708090100

Above FBBelow FB

very fine

fine

fine loamy fine silty

coarse loamy coarse siltysandy

Fig. 4. Particle size distribution and textural classification of climosequence microlow profile horizons.


microhighs and bowl-like microlows (Lynn and Williams, 1992). Adjacent microlows and

microhighs typically have similar, but not identical, bulk geochemistry and stable isotope

compositions (Stiles, 2001; Miller, 2000) and the depths to the functional boundary in

adjacent microlows and microhighs indicate somewhat contrasting physical and geo-

chemical behavior within the Vertisol microtopographic domains (data not shown here: see

Stiles, 2001). Microhighs are often complicated structurally by thrusting, are better drained

and tend to be depleted in organic matter and enriched in secondary evaporites compared

to adjacent microlows. Microlows have previously been demonstrated to behave as more

physically and chemically closed systems (Wilding et al., 1990; Driese et al., 2000).

Bulk geochemical analysis revealed notable changes in the Ti/Zr volumetric ratio at

depth in each pedon (Fig. 2). Tukey’s w procedure using Studentized range and pair-wise

comparisons of the microlow Ti/Zr values were used to statistically determine the

functional boundary depth between the upper and lower units (Table 2). Mean volumetric

contents of the two elements show that Ti contents do not vary greatly between upper and

lower units (� 4.7% to � 31.6%), confirmed by the mean separation and least significant

difference (LSD) values shown in Table 2. Only the wettest Laewest profile (LAW 239)

showed significance at P < 0.025. By comparison, Zr differences vary from � 75.5% to

� 239.3% between the upper and lower units and are all statistically significant at

P < 0.001 (Table 2).

Fig. 5. Photomicrographs of textural contrasts across the functional boundary in two profiles.


The depth and magnitude of the volumetric Ti:Zr shifts differ in each pedon and cannot

be strictly ascribed to climate (i.e., MAP), because the shallowest inflections occur within

soils formed in the intermediate precipitation range (1000–1200 mm MAP) of the

climosequence. These considerable and pervasive differences could be caused either by:

(1) a regional sedimentological discontinuity, or (2) differences resulting from pedogenic

processes. To determine which component more greatly affects the Ti/Zr ratios, element

versus ratio trends were examined. Titanium (Fig. 3A) is relatively insensitive to increases

in the Ti:Zr ratio, even across the functional boundary. In contrast, volumetric Zr content

shows strong correlation with the Zr:Ti ratios (R = 0.926, P < 0.001) for all pedons (Fig.

3B). Volumetric Zr contents in the lower soil domains (below the functional boundary) are

less than those of the upper units. These results indicate that variability in Ti:Zr volumetric

ratios are largely influenced by Zr contents, with Ti behaving more conservatively.

3.2. Particle size distribution and component comparisons

The Vertisol pedons within the climosequence are fine-textured, with clay contents of

35.8–72.6 wt.% (Fig. 4). The sand and coarse silt (SCST) size fractions of these fine-

textured soils contribute significantly to the observed compositional trends and are closely

related to the position of the functional boundary. Sand and silt contents are highest in the

Victoria and drier Laewest pedons (Fig. 1). In all profiles, soil textures tend to coarsen

upward and are significantly finer below the functional boundary (Fig. 4, filled symbols).

Fig. 6. Mean Ti and Zr composition of size fraction separates from LAC 481 and LAW 239 microlow profiles.


Micromorphologic differences across the boundary are pronounced between the upper and

lower units (Fig. 5). Photomicrographs show more abundant and coarser-textured quartz

sand grains above the boundary in both profiles (Fig. 5A,C). Similarly, there are

significantly fewer large grains in the lower portions of these profiles (Fig. 5B,D), and

most of the grains are located along voids within the matrix.

Geochemical analyses of the size fractions (Fig. 6) show that Ti tends to be enriched in

the finer fractions ( < 20 Am; Fig. 6A) while Zr is enriched in the coarse silt (20–50 Am),

and is particularly concentrated above the functional boundary in LAW 239 (Fig. 6B).

Trends in LAC 481 are similar to those noted in LAW 239, although intrinsic composi-

tional values are slightly different. Volumetric Zr contents show significant correlation to

the SCST content in each soil series within the climosequence (Fig. 7). Regression

Fig. 7. Separated regression analyses of sand + coarse silt (SCSi) content versus volumetric Zr content by soil

series.


analyses of each soil series show positive correlation of volumetric Zr contents and wt.%

SCST, with r2 values from 0.69 to 0.84 (P < 0.001). The slopes of the series regressions

are steeper in series having relatively low MAP (i.e., Laewest and Victoria series),

indicating a significant increase of Zr with even minor additions to the SCST fraction.

Titanium does not correlate significantly to the SCST content (R =� 0.20; r2 = 0.04,

P < 0.001; Fig. 8A), however, the relationship of Zr with SCST (a combined regression of

all data illustrated in Fig. 7) is significant (r2 = 0.76; P < 0.001; Fig. 8B). Approximately

0.10 wt.% of the sand component is estimated to be zircon, and the contribution of fine

fraction Zr to the regression relationship is negligible (X intercept = 2.41�10� 3). A

greater proportion of coarse-sized material occurs in the upper portion of the profiles,

0 10 20 30 40

0.0

0.5

1.0

1.5

2.0

2.5

Zr (mg cm-3) = 0.05 SCSi wt %

5

7

9

11

13

15

A.

B.

Sand and coarse silt content (wt %)

Ti (mg cm-3) = 9.06 - 0.03 SCSi wt %

r2 = 0.76***

Zr

(mg

cm-3

)T

i (m

g cm

-3)

r2 = 0.04ns

Fig. 8. Linear regression relationship between sand and coarse silt (SCSi) content and volumetric elemental

contents in all climosequence microlow profiles: (A) Ti; (B) Zr.


suggesting that 20–50 Am zircons are relatively abundant in the upper portion of the

weathering profiles in the climosequence. Lack of correlation between Ti content and

SCST indicates that Ti is rare in the >20 Am fraction, or may have been preferentially

removed. Relatively large amounts of Ti in samples with no SCST (Fig. 8A) further

indicates that most of the Ti occurs in finer texture components.

3.3. Qualitative analysis of surface microfeatures

To further elucidate processes affecting the distribution of Ti and Zr within the profiles,

surface features of rutile and zircon grains were assessed using scanning electron

microscopy (SEM). The features were categorized as either being influenced by: (1)

Fig. 9. Photomicrographs of characteristic very fine sand-sized rutile and zircon grains.

Table 3

Percent values of SEM survey characteristics for rutiles and zircons from units above and below the functional

boundary in climosequence Vertisol pedons

Pedon (MAP in mm) Hydrogeochemical Physical transport

Rutiles Zircons Differencea Rutiles Zircons Difference

%

LEG 245A (1437) 91/91b 36/5 55/86 55/57 62/57 � 7/0

LAC 201 (1321) 87/84 22/13 55/71 49/53 69/67 � 20/� 14

LAC 481 (1124) 99/93 13/22 86/71 56/50 80/69 � 24/� 19

LAW 469 (1066) 82/85 17/24 65/61 46/49 58/60 � 12/� 11

LAW 391 (924) 94/97 15/13 79/84 50/50 70/73 � 20/� 23

VIC 409 (844) 96/58 31/30 64/28 63/54 83/67 � 20/� 13

Overall means 92/85 22/18 70/67 53/52 71/65 � 18/� 13

a Categorical percentage differences between features expressed in rutile grain morphology versus zircon

grain morphology. Positive values indicate rutiles had higher percentage of expression than zircons, negative

values indicate zircons had a higher percentage of expression than rutiles.b Mean categorical percentages above FB/below FB.


physical processing (i.e., grains show percussion marks, hairline cracks, abrasion, grain

frosting and rounding) suggestive of transport, or (2) weathering/chemical degradation

(i.e., grains show dissolution at edges or etch pits and surface roughness). Rutiles

predominantly show evidence of weathering (80–100% of grains in the upper unit,

60–95% in the lower unit; Fig. 9B,F; Table 3), whereas zircon morphology mostly

indicates physical damage by transport reworking (60–85% of grains in the upper unit,

55–75% in the lower unit; Fig. 9A,C; Table 3). Grain-frosting and rounding, which attest

to long-term aeolian transport, are evident in the driest profiles (Victoria and dry Laewest

series, Fig. 9C,D), particularly in the upper unit. Rutile grains in these profiles shows

markedly less evidence of hydrogeochemical weathering (Fig. 9: compare D to B, F). The

presence of occasional euhedral zircons in the upper unit of alluvially sourced pedons,

such as in Lake Charles 481 (Fig. 9E), strongly suggests external additions from

volcanogenic lithologies to the south–southwest. Analysis of variance for the pooled

categories (transport versus chemical features) between very-fine sand-sized rutile and

zircon is highly significant (PV 0.0001), regardless of MAP. Rutile grains always show

greater evidence of chemical weathering, to which zircon is less susceptible, whereas

zircons always show more evidence of transport than rutiles (Table 3).

4. Discussion

What is causing the pronounced shifts in Ti and Zr content with depth and the

contrasting weathering of rutile and zircon grains in the profiles? Do differences in the

morphological and micromorphological characteristics of soil occurring above and

below the functional boundary indicate a lithologic discontinuity, a regional scale

depositional event that preceded pedogenesis of the upper unit, or a change in

physicochemical pedogenic inputs? A difference in depositional units can be largely

discounted by the geomorphology and sedimentology of the setting. As the Gulf of


Mexico Basin subsides under an increasing sediment load, the marginal fluvio-deltaic

deposits that blanket the Texas Gulf Coast may be depositing coarser material over the

finer-textured (clayey) deposits as part of a complex glacio-eustatic retrogradational

sequence. However, the great thickness of the upper units in the Vertisol profiles (1.6–

2.6 m), the relatively advanced morphological maturity, and temporal limitations make

a major depositional change an unlikely scenario in this relatively low-energy

depositional setting.

4.1. Discerning pedogenic processes

Vertisols along the climosequence are exposed to different levels of physical and

geochemical pedogenic processes and the functional boundary may indicate the relative

influence and intensities of these processes. Vertisols are remarkable for their expres-

sions of physical processing, with characteristic slickensides and microtopographic

differentiation generated and maintained by seasonal shrink–swell mechanisms (Lynn

and Williams, 1992). Seasonal cracking, which occurs more frequently within micro-

lows, creates a solum responsive to meteoric inputs down to the depth of cracking

(Wilding et al., 1990). Microhighs tend to act not only as foci for upward physical

material transport, but the relatively high relief of microhighs may allow them to

physically ‘‘shed’’ materials into the bowl-like depressions of the microlow. Such

microtopographic relief reaches maximum expression in climatic zones experiencing

large seasonal soil-moisture deficits and is relatively subdued in both the wetter and

drier ends of the climosequence (Stiles, 2001).

In wet climates with lower seasonal moisture deficits (i.e., receiving greater amounts of

and more constant delivery of precipitation throughout the year), Vertisol pedogenesis is

strongly influenced by relatively high biotic activity/productivity. Within the climose-

quence, Vertisols with MAP >1300 mm have evidence of extensive bioturbation,

particularly crayfish (Fallicambarus devastator) krotovina. Krotovina extended as deep

as 2 m in the profiles and accumulation of crayfish wastes at the surface was notable in

microlows. High productivity enhances bioturbation and overall elemental cycling,

ameliorating the strictly physical effects of seasonal shrink–swell processes. This assess-

ment was reinforced by strong relationships of mass-balance translocation trends of

biocycled elements such as calcium, magnesium, and potassium with MAP (data not

shown; Stiles, 2001).

Vertisols formed in lower MAP climates with extended periods of soil moisture deficits

have limited slickenside and microtopographic development simply because there is not

enough available soil water, even in the wet season, to generate the necessary clay-

swelling and associated physical forces (Wilding and Tessier, 1988). Cracks form and

remain open for longer periods of time during dry periods, providing ample conduits for

wind-borne dust into the profiles. Aeolian accumulations tend to form a blanket that

thickens over time in most soils (Yaalon, 1987) and are effectively intercalated into

Vertisol cracks. Lower soil moisture also effectively retards mineral weathering, lessens

active transport of coarse fragments, and contributes to sand preservation in the profile.

This argument is supported by the stronger relationship of Zr with SCST (e.g., steeper

slopes of the linear correlations) as MAP decreases (Fig. 7).


4.2. Conceptual model

Changes in the functional boundary depth show the contrasting intensities of pedogenic

processes and expression within MAP zones. We hypothesize that the predominance of

either (1) physical processing and aeolian addition of material, or (2) hydro/biogeochem-

ical activity influence the functional boundary depth in the climosequence profiles (Fig.

10). On the basis on zircon morphology, particle size distribution and Ti/Zr ratios, pedons

formed under low MAP are interpreted to be dominated by physical pedogenic processes

and aeolian addition of material to the soil during peodgenesis. These pedons show the

deepest expression of the functional boundary. By comparison, pedogenesis in the highest

MAP regimes is dominated by hydro/biogeochemical activity. In these pedons, the

functional boundaries are expressed at higher levels in the soil profile. Vertisols formed

under intermediate MAP are subject to a combination of both and have functional

boundaries similar to, or slightly greater than the wetter pedons.

More notable are differences in the depth of the functional boundaries in corresponding

microhigh and microlow pedons (Fig. 10). These are similar to the microlow at either

extreme of the climosequence, yet very different in intermediate MAP pedons. In

Vertisols, mechanical deformation induced by shrink–swell actions of the clays and gilgai

microtopography is controlled by available water over the seasons. Thus, the functional

Fig. 10. Conceptual model of climosequence Vertisol pedogenesis based on depth to functional boundary as a

function of MAP.


boundary depth differences between microhighs and microlows from dry to intermediate

to wet MAP Vertisols can be partially explained by morphological variations and

mechanical actions within the profiles. In low MAP Vertisols, gilgai development is

modest (Stiles, 2001). The microlows act as catchment basins for material brought in by

wind or shed from microhighs. Microlows in low MAP Vertisols more effectively capture

the limited rainfall and are therefore ‘‘wetter’’ than corresponding microhighs, resulting in

deeper cracking during the dry season, and a functional boundary in the microlow that is

deeper than the microhigh (Fig. 10). At the middle of the MAP scale, gilgai are most

strongly developed (Stiles, 2001), indicating the importance of physical translocation and

mechanical deformation induced by clay shrink–swell. This, combined with attenuated,

but still active, hydro/biogeochemical activity results in shallow and irregular functional

boundary depths (Figs. 2 and 10). Although these profiles have experienced some level of

aeolian input (Fig. 9E), degradation/eluviation rates appear equivalent to accumulation

rates.

Within the high MAP Vertisols, water is not limiting and the depth of the functional

boundary increases in response to hydro/biogeochemical mechanisms. The presence of

abundant dissolved organic substances contributes to carbonate and salt solubility and

creates favorable conditions for lessivage. At the same time, bioproductivity increases,

producing a positive feedback cycle where recycled material from deeper in the profile is

moved to the surface, producing more abundant organic matter that, in turn, enhances

geochemical transport (Stiles, 2001). Microlows generally have higher organic contents

(Coulombe et al., 1996), and also tend to be habitats for diverse communities of soil-

dwelling fauna. Physical transport of materials to the surface via meso- and macrofauna

(particularly crayfish) tends to disrupt hydrogeochemical trends and attenuates strict

downward-weathering signatures. Thus, at high MAP regimes, the depth to the functional

boundary is shallower in the microlow than the microhigh, where crayfish accumulations

are less subject to intense reworking and accumulation rates slightly exceed erosion.

4.3. Titanium as a mass-balance strain indicator in clay-rich soils

A principle objective for this work is to evaluate which ‘‘immobile’’ element is best to

use for mass-balance analysis in a Vertisol climosequence. Geochemical mass-balance

calculations require identification of an immobile index element to determine residual

enrichments and translocations. Soil strain, defined as:

ei;w ¼ ðqpCi;p=qwCi;wÞ � 1

where ei,w is strain in weathered material w based on immobile element i, q is the soil bulk

density, p is unweathered parent material interval, and C is concentration of immobile

element i in p or w, strain is the basis by which other elements translocations are

determined in mass-balance relationships (Brimhall and Dietrich, 1987).

No element is truly immobile during intense pedogenesis, but only relatively immobile

due to retarded hydrogeochemical or physical processing. Thus, the choice of an immobile

strain element is dependent upon the pedogenic setting. Elements affected by influxes of

exotic material or fluids, such as the Zr–SCST relationship, can only be used when such


accumulations are dynamically balanced with internal processes. Such reasoning also

holds true with elements found in minerals susceptible to hydrogeochemical alteration

such as Ti in rutile. If the pedogenic conditions are such that distribution of the element

does not change, although there may be some redistribution among primary and secondary

phases (i.e., Vertisol fine-fractions), then that element is the strain indicator of choice in

that particular setting. Another reason for using Ti is analytical: XRF-detectable Ti is

always much more abundant than Zr in these soils. Titanium contents are 3–40 times

higher than Zr, with associated instrumental errors of 0.85–2.38% for TiO2, but as much

as 1.85–25.00% for Zr. When used in Vertisol mass-balance calculations, Ti is the

conservative element of choice allowing for maximum depiction of the pedogenic

responses of an element of interest, rather than its relationship to the strain indicator itself.

The mean strain (ei) for both Ti and Zr in all microlow/microhigh pedons in the

climosequence is strongly correlated (r2 = 0.83, P < 0.001; Fig. 11), suggesting both

elements reflect overall similar similarity in their pedogenic behavior. The regression line

also shows that eZr values are overall approximately four times greater than corresponding

eTi, due primarily to Zr accumulations in the upper unit of most of the pedons. Strain

calculated from Ti shows the same trends as eZr, but with lower magnitude suggesting

more conservative behavior of Ti within the Vertisol profiles. Mean eZr for depth intervals

ranged from � 0.68 to + 0.20 (i.e., � 68% to + 20%) in microlows, and � 0.60 to + 0.47

in microhighs. By comparison, eTi ranged from � 0.15 to + 0.05 in microlows, and

Strain (εTi)

-0.2 -0.1 0.0 0.1 0.2

Stra

in (

ε Zr)

-0.75

-0.50

-0.25

0.00

0.25

0.50

0.75Strain (εZr) = -0.158 + 3.972 Strain (εTι)r2 = 0.83***

Fig. 11. Linear regression of eTi and eZr calculation for mean depth interval for all profiles.


� 0.09 to + 0.18 in microhighs, indicating markedly less pronounced volume loss/gain.

The conservative nature of Ti in fine-textured Vertisols makes it a more suitable strain

indicator than Zr, and an indicator that helps quantify pedogenic effects more reliably,

both within and between Vertisol pedons. Thus, it is important to assess the immobility

or internal variations are due entirely to weathering of the parent sediment or are there,

in fact, external factors that may skew the trends.

5. Conclusions

Geochemical and physical evidence delineate a functional boundary in Vertisol pedons,

which indicates a shift in primary pedogenic intensity. Depth to the functional boundary

indicates the relative influence of contrasting pedogenic mechanisms (hydrogeochemical

versus physical) across the climosequence. The functional boundary defines the depth of

the superactive pedogenesis domain in the Vertisols, where external fluxes create a

dynamic open-system. Microscopic (SEM) survey of grain morphology revealed rutile

grains were more weathered relative to zircon, whereas zircons always showed more

physical grain damage indicative of extensive transport. This relationship was expressed

throughout the Vertisol profiles, regardless of the depth of the functional boundary,

although it was slightly more pronounced in zircon features as a function of MAP. There is

typically residual enrichment of Zr and Ti in the upper portions of Vertisols that results

from loss of more mobile elements during pedogenesis. However, the strong correlation of

Zr with the small, but significant, sand and coarse silt fraction in these clay-rich soils,

along with the SEM evidence that many very fine sand-sized zircon grains have been

extensively transported from outside areas, argues that Zr is predominantly an indicator of

physical processes and aeolian input rather than overall weathering during Vertisol

pedogenesis. Despite evidence of minor weathering, Ti is conserved within weathering

clay-rich profiles and is better suited as a closed-system strain indicator to determine

mobile elemental translocation intensity using mass-balance.

The concept of a ‘‘lithologic discontinuity’’ as it applies to U.S. Soil Taxonomy is not

fully defined in terms of pedogenesis, although there are several criteria given by the Soil

Survey for its recognition (Soil Survey Staff, 1998). In some cases, there are true breaks in

the lithologic composition of profiles, but in many cases, lithologic discontinuities may, in

fact, be the translocation crossover described by Brimhall et al. (1991) or the functional

boundary described in this paper. Within the Vertisol climosequence, there are only slight

variations in the initial parent material in which soil begins to form, but over time, open-

system dynamics create a complex lithology characteristic of the soil itself. The depth to

the functional boundary becomes an interpretive tool for Vertisol pedogenesis using both

Ti and Zr as indicators of contrasting pedogenic mechanisms. Within the Texas Gulf Coast

climosequence, the interpretations gained from using this tool are:

(1) All pedons experienced some level of aeolian input;

(2) Hydro/biogeochemical processes dominate wet Vertisols, subduing expression of

shrink–swell phenomenon and slickenside development and microtopographic

differences;


(3) Physical processes dominate the pedogenesis of Vertisols formed in dry climates,

where wind-borne material infiltrates into long-lived seasonal cracks and is only

slowly weathered;

(4) Mid-range MAP Vertisols show the maximum additive interaction between the

physical and hydrogeochemical pedogenic processes and the most extreme differences

in the physical and chemical expression of microtopographic highs and lows.

Acknowledgements

This project was carried out in conjunction with a U.S. Department of Agriculture-

NRCS reevaluation of the Texas Gulf Coast Prairie soil resource area. We are grateful to

NRCS district soil scientists for logistical field support, particularly in selection of

sampling sites. We thank Amy Robinson and Ellen Robey (UT-Knoxville) for sample

preparation and analysis, and Richard Drees and Larry Wilding of Texas A&M University

for additional data and helpful discussion.

References

Bernard, H.A., LeBlanc, R.J., 1965. Resume of the Quaternary geology of the northwestern Gulf of Mexico

province. In: Wright, H.E., Frey, D.G. (Eds.), The Quaternary of the United States. Princeton Univ. Press,

Princeton, NJ, pp. 137–185.

Birdseye, H.A., Aronow, S., 1991. New evidence for a young late Wisconsin age for the Prairie Formation,

Texas, USA. Geol. Soc. Am., Abstr. Program 23, A223.

Birkeland, P.W., 1999. Soils and Geomorphology, 3rd ed. Oxford Univ. Press, New York.

Blake, G.R., Hartge, K.H., 1986. Bulk density. In: Klute, A. (Ed.), Methods of Soil Analysis: Pt. 1. Physical and

Mineralogical Methods, 2nd ed., vol. 9. Soil Sci. Soc. Amer. Agron. Monogr., pp. 363–375.

Blum, M.D., Price, D.M., 1994. Glacio-eustatic and climatic controls on Quaternary alluvial plain deposition:

Texas Coastal Plain. Gulf Coast Assoc. Geol. Soc. Trans. 44, 85–92.

Brimhall, G.H., Dietrich, W.E., 1987. Constitutive mass balance relations between chemical composition, vol-

ume, density, porosity, and strain in metasomatic hydrochemical systems: results on weathering and pedo-

genesis. Geochim. Cosmochim. Acta 51, 567–587.

Brimhall, G.H., Chadwick, O.A., Lewis, C.J., Compston, W., Williams, I.S., Danti, K.J., Dietrich, W.E., Power,

M.E., Hendricks, D., Bratt, J., 1991. Deformational mass transport and invasive processes in soil evolution.

Science 255, 695–702.

Coulombe, C.E., Dixon, J.B., Wilding, L.P., 1996. Mineralogy and chemistry of Vertisols. In: Ahmad, N.,

Mermut, A. (Eds.), Vertisols and Technologies for their Management: Developments in Soil Science, vol.

24. Elsevier, New York, pp. 115–200.

Darmody, R.G., 1985. Weathering assessment of quartz grains: a semi-quantitative approach. Soil Sci. Soc. Am.

J. 49, 1322–1324.

Driese, D.G., Mora, C.I., Stiles, C.A., Joeckel, R.M., Nordt, L.C., 2000. Mass-balance reconstruction of a modern

Vertisol: implications for interpreting the geochemistry and burial alteration of paleo-Vertisols. Geoderma 95,

179–204.

Fitzpatrick, R.W., Chittleborough, D.J., 2002. Titanium and zirconium minerals. In: Dixon, J.B., Schulze,

D.G. (Eds.), Soil Mineralogy with Environmental Applications. Soil Sci. Soc. Amer. Book Series, vol. 7,

pp. 667–690.

Gee, G.W., Bauder, J.W., 1986. Particle-size analysis. In: Klute, A. (Ed.), Methods of Soil Analysis: Pt. 1.

Physical and Mineralogical Methods, 2nd ed., vol. 9. Soil Sci. Soc. Amer. Agron. Monogr., pp. 383–411.


Jackson, M.L., 1985. Mineral fractionation for soils. In: Jackson, M.L. (Ed.), Soil Chemical Analysis, 2nd ed.,

pp. 101–166. Published by author, Madison, WI.

Jackson, M.L., Sherman, G.D., 1953. Chemical weathering of minerals in soils. Adv. Agron. 5, 219–318.

Kaup, B.S., Carter, B.J., 1987. Determining Ti source and distribution within a Paleustalf by micromorphology,

submicroscopy and elemental analysis. Geoderma 40, 141–156.

Lynn, W.C., Williams, D., 1992. The making of a Vertisol. Soil Surv. Horiz. 33, 45–50.

Marshall, C.E., Haseman, J.F., 1942. The quantitative evaluation of soil formation and development by heavy

mineral studies: a Grundy silt loam profile. Soil Sci. Soc. Am. Proc. 7, 448–453.

Miller, D.L., 2000. Occurrence and stable isotope compositions of soil carbonate and organic matter within a

climatic transect of modern Vertisols along the costal prairie of Texas. MS thesis, The University of Tennes-

see, Knoxville, TN.

Reheis, M.C., 1990. Influence of climate and eolian dust on the major-element chemistry and clay mineralogy of

soils in the northern Bighorn Basin, U.S.A. Catena 17, 219–248.

Singer, M.J., Janitsky, P., 1986. Field and laboratory procedures used in a soil chronosequence study. U.S. Geol.

Surv. Bull. 1648 (16 pp.).

Soil Survey Staff, 1995. Soil Survey Laboratory Information Manual. Soil Survey Invest. Report No. 45, v. 1.0.

U.S. Government Printing Office, Washington, DC. 305 pp.

Soil Survey Staff, P., 1998. Keys to Soil Taxonomy, 8th ed. U.S. Government Printing Office, Washington, DC.

326 pp.

Stiles, C.A., 2001. Pedogenic relationships in a Texas Vertisol climosequence defined by geochemical mass

balance of whole soil and chemistry of iron–manganese nodules. PhD dissertation, The University of

Tennessee, Knoxville, TN.

Tejan-Kella, M.S., Fitzpatrick, R.W., Chittleborough, D.J., 1990. Scanning electron microscope study of zircons

and rutiles from a podzol chronosequence at Cooloola, Queensland, Australia. Catena 18, 11–30.

Wilding, L.P., Tessier, D., 1988. Genesis of Vertisols: shrink–swell phenomena. In: Wilding, L.P., Puentes, R.

(Eds.), Vertisols: Their Distribution, Properties, Classification, and Management. Texas A&M Univ. Print.

Center, College Station, TX, pp. 23–57.

Wilding, L.P., Williams, D., Miller, W., Cook, T., Eswaran, H., 1990. Close interval spatial variability of

Vertisols: a case study in Texas. In: Kimble, J.M. (Ed.), Proceedings of the 6th International Soil Correlation

Meeting (VI ISCOM)—Characterization, Classification, Utilization of Cold Aridisols and Vertisols. USDA

Soil Cons. Serv., Nat. Soil Surv. Center, Lincoln, pp. 232–247.

Yaalon, D.H., 1987. Saharan dust and desert loess: effect on surrounding soils. J. Afr. Earth Sci. 6, 569–571.

Pedogenic processes and domain boundaries in a Vertisol climosequence: evidence from titanium and zirconium distribution and morphology

Documents