See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233150113

Serpentine Soil Redness, Differences among Peridotite and

Serpentinite Materials, Klamath Mountains, California

Article in International Geology Review · July 2010

DOI: 10.2747/0020-6814.46.8.754

CITATIONS READS

56 687

1 author:

Earl Alexander
Soils and Geoecology
40 PUBLICATIONS 830 CITATIONS

SEE PROFILE

All content following this page was uploaded by Earl Alexander on 28 December 2017.

The user has requested enhancement of the downloaded file.

International Geology Review, Vol. 46, 2004, p. XXX–XXX.
Copyright © 2004 by V. H. Winston & Son, Inc. All rights reserved.

Serpentine Soil Redness, Differences among Peridotite

and Serpentinite Materials, Klamath Mountains, California
E. B. ALEXANDER1
Soils and Geoecology, 1714 Kasba Street, Concord, California 94518-3305

Abstract

Peridotite soils were perceived to be redder than serpentinite soils. These redness differences
were confirmed by relating soil redness to bedrock specific gravities and to heavy mineral concen-
trations in fine sand fractions of soils. The redness differences are explained by mineralogical dif-
ferences between peridotite and serpentinite. Soil redness in well-drained soils of the Klamath
Mountains is closely related to free iron in oxides and oxyhydroxides. Most of the iron in peridotite
is in olivine, which is readily weathered to release iron that is then oxidized in well-drained soils,
whereas most of the iron in serpentinite is in magnetite and serpentine. Magnetite is very resistant
to weathering and serpentine weathers more slowly than olivine and pyroxene. This relationship
predicts untested redness differences in soils derived from serpentinized dunite (redder soils) com-
pared to those from more silicic serpentinized peridotite such as lherzolite.

Introduction serpentinites generally have grey weathering sur-

faces.
SOILS WITH EITHER peridotite or serpentinite parent Serpentine soils most common in the Klamath
materials are commonly called serpentine soils. Mountains are Xerepts, Cryepts, Xerolls, Cryolls,
Peridotite and serpentinite have practically the and Xeralfs. Less common are Udepts, Udolls,
same chemical composition, except that serpen- Aquolls, Xerults, and Xererts. Only in the southern
tinite has about 13% (130 g/kg) more water (Cole- exposure of the Rattlesnake Creek terrane (Fig. 1)
man, 1971). Nevertheless, weathered surfaces on have serpentine soils been mapped in detail. All of
peridotite generally look quite different from weath- the more common serpentine soils in the Klamath
ered surfaces on serpentinite, and the soils derived Mountains are well represented in the southern
from these rocks are generally different. exposure of the Rattlesnake Creek terrane, except
Peridotite exposed on ground surfaces generally for Cryepts and Cryolls, and none of the soils there
weathers to yellowish brown (or reddish brown after have udic soil moisture regimes (Soil Survey Staff,
heating in wildfires) and weathered serpentinite 1999). The most striking difference among serpen-
rock surfaces are generally light grey to white. tine soils in the Rattlesnake Creek terrane is that
Weathering surfaces of harzburgite, the dominant Xeralfs are dominant on peridotite and Xerolls are
variety of peridotite in the Klamath Mountains, dominant on serpentinite (Alexander, 2003). A pre-
develop a warty appearance with pyroxene crystals liminary investigation indicated that soils on peri-
projecting a few millimeters above the olivine, dotite are redder than those on serpentinite
which weathers much more rapidly than pyroxenes. (Alexander, 2000). Additional data from the Rattle-
Dunite lacking pyroxene develops smooth, yellow- snake Creek terrane, the Trinity ultramafic body,
ish brown surfaces with a “buckskin” appearance. and from Bluff Lake ridges in the east-central
Also, some massive serpentinite, in contrast to Klamath Mountains (Fig. 1) were analyzed to exam-
sheared serpentinite, may have yellowish brown ine further the hypothesis that soils derived from
weathering surfaces. Serpentinite with yellowish peridotite are redder than those derived from
brown weathering surfaces may be peculiar to that serpentinite. Results of these data are presented
formed by the alteration of dunite, because massive here, with an explanation of differences in redness
from soils in peridotite compared to those in
1Email: [email protected] serpentinite.

0020-6814/04/XXX/XXX-11 $25.00 1
2 E. B. ALEXANDER

FIG. 1. Locations of the southern exposure of the Rattlesnake Creek terrane (RCT), the Trinity ultramafic body, and
the Bluff Lake area in the Klamath Mountains based on a geologic map (Irwin, 1994).

Soils were described according to terminology in

Serpentine Soils of the the Soil Survey Manual (Soil Survey Staff, 1993).
Rattlesnake Creek Terrane Only shallow (18 to 50 cm deep) and moderately
deep soils (50 to 100 cm to bedrock) with argillic
A survey of serpentine soils in the southern horizons were included in redness comparisons—a
exposure of the Rattlesnake Creek terrane included total of 126 pedons. The reddest colors (moist) in
the description of 196 pedons (Alexander, 2003). argillic horizons, generally the uppermost Bt subho-
Nearly two-thirds of these were in Alfisols or Mol- rizons, were utilized for comparisons. Munsell col-
lisols between 18 to 100 cm deep over either peri- ors were converted to redness ratings (Rr) by
dotite or serpentinite bedrock. Nearly all of the subtracting the prefix to YR from 15 (or 15 minus
serpentine soils with argillic horizons are in this 12.5 for 2.5Y), multiplying by the chroma, and
depth range, except those in landslides. dividing by the value (Alexander, 2000).
Methods Results
Serpentine soil parent materials were separated Of 126 soils in the redness comparison, 50 are on
into partially and completely serpentinized classes, partially serpentinized peridotite, or simply “peri-
based on presence or absence of pyroxenite grains dotite,” and 76 are on completely serpentinized
identified in hand specimen. Apparently any dunite peridotite, or serpentinite. They are mostly Haplox-
that might have been present at the pedon sites has eralfs and Argixerolls with mesic soil temperature
been completely serpentinized, because all of the regimes, and a few with thermic and cryic regimes
observed rocks that contained olivine also had (Soil Survey Staff, 1999). Redness differences from
pyroxenite. shallow to moderately deep soils were insignificant
 SERPENTINE SOIL REDNESS 3

FIG. 2. Distributions of serpentine soils in the Rattlesnake Creek terrane (126 sample pedons) by redness rating
classes: low (Rr < 5.1), moderate (Rr = 5.1–10), moderately high (Rr = 10.1–15), and high (Rr > 15).

for both peridotite (unpaired t = 0.877, 48 d.f.) and to other soil properties. Soils at the 22 sites are
serpentinite (unpaired t = 0.497, 74 d.f.). Xerepts, Xerolls, and Xeralfs with mesic and frigid
Distributions of soils in four classes of redness soil temperature regimes (Soil Survey Staff, 1999).
(Fig. 2) show distinct differences between soils on
peridotite and those on serpentinite. Color hues Methods
range from 2.5YR to 2.5Y. Moist colors representa- Soils were sampled at 0 to 10, 10 to 30, and 30 to
tive of the low (Rr < 5.1), moderate (Rr = 5.1–10), 60 cm depths. Soil colors were ascertained by com-
moderately high (Rr = 10.1–15), and high (Rr > 15) parison to colors on a Munsell chart. They were con-
classes are 10YR 4/3, 7.5YR 4/4, 5YR 4/6, and verted to redness ratings by the same algorithm
2.5YR 4/6. Because the redness rating distributions utilized for soils of the Rattlesnake Creek terrane.
are not even close to normal (bell-shaped) distribu- After dry sieving to remove particles > 2 mm, Fe was
tions, means are practically meaningless. The extracted in solutions of Na-dithionite and Na-cit-
majority of soils on peridotite are in the moderately rate, and Fe was ascertained by atomic absorption
high class, and many are in the high class. The photometry (Alexander et al., 1989). For samples
majority of soils on serpentinite are in the low class, from the 30 to 60 cm depth, fine sand (75–250 µm)
and only one is in the high class. Soils on peridotite was separated from the fraction <2 mm by dry siev-
definitely tend to be redder than those on serpen- ing, and light grains were separated from the fine
tinite in southern exposures of the Rattlesnake sand by flotation on bromoform, SG = 2.89 (Alex-
Creek terrane. ander et al., 1990). Magnetic grains were separated
from the heavy fraction with a hand magnet.
Serpentine Soils of the
Results
Trinity Ultramafic Body
Redness ratings were not correlated significantly
The Trinity ultramafic body, probably an ophio- with specific gravities of rocks from the soil parent
lite, in the east-central Klamath Mountains, is the materials, presumably because of rock weathering.
largest in western North America. Soils were sam- Extractable Fe from the 10 to 30 cm depth was
pled at 22 sites widely distributed across the body to related to the mass of the heavy (SG > 2.89) and the
examine forest productivity on serpentine soils nonmagnetic heavy fractions of fine sand from the
(Alexander et al., 1989). Data from those 22 sites 30 to 60 cm depth (Fig. 3), and redness ratings were
are helpful in examining relationships of soil color related to extractable Fe concentrations r = 0.675,
4 E. B. ALEXANDER

FIG. 3. Relations of citrate-dithionite extractable iron from the 10 to 30 cm depths to mass of the heavy (SG > 2.89)
fractions in fine sand from 30 to 60 cm depths in 22 soils of the Trinity ultramafic body.

FIG. 4. Relations of redness rating (moist) of samples from the 10 to 30 cm depths to mass of the heavy nonmagnetic
fraction (SG > 2.89) in fine sand from 30 to 60 cm depths in 22 soils of the Trinity ultramafic body.

n = 22). Then, redness ratings for samples from 10 weathered than serpentine and the felsic minerals,
to 30 cm depths were presumed to be related to the and they contain more Fe; therefore, they are the
mass of the nonmagnetic heavy fractions of fine sand major sources of Fe in these soils.
(Fig. 4).
Minerals in the nonmagnetic heavy fractions of Serpentine Soils on Bluff Lake Ridges
fine sand were mostly olivine, pyroxenes, and
amphiboles, and serpentine dominated the light The Bluff Lake ridges are linear features with
fractions in all samples, except for two that con- serpentinite diamicton on the southern ends of each
tained quartz and two others that contained feld- and peridotite diamicton on the northern ends (Fig.
spars. The heavy mafic minerals are more readily 5). They are in the shadow of Cement Bluff, which is
 SERPENTINE SOIL REDNESS 5

FIG. 5. Topographic and surficial geology map showing the Bluff Lake ridges and Cement Bluff. Legend: A = recent
alluvium; C = recent colluvium; Dp = Bluff Lake peridotite diamicton; Ds = Bluff Lake serpentinite diamicton; G = High
Camp glacial drift, Wisconsin till; M = Cement Bluff glacial drift, pre-Wisconsin till; P = peridotite bedrock. Contours:
solid = 40-foot interval, above sea level; dashed = 3-meter interval, from lake level.

over 100 meters high. Predominantly serpentinite 3) on the south. The extent of the serpentinite diam-
glacial drift is exposed in the lower part of the bluff icton was easy to map, but the peridotite diamicton
and predominantly peridotite glacial drift in the was more difficult to differentiate from older (Wis-
upper part. These strata are the apparent contribu- consin) glacial drift over which the diamicton
tors of the diamicton that is in the Bluff Lake ridges, appears to have been deposited, at least on the distal
but it is not clear how it was transported to the ridges ridge (Fig. 5).
and how the segregation of serpentinite and peridot- Two possible modes of diamicton transport to the
ite materials was maintained in transport to the ridges are (1) by glaciers or (2) by gravitational
ridges. The boundaries between peridotite and ser- transport over ice or snow. Thus, the ridges may be
pentinite diamictons are rather distinct, with no either glacial moraines or pronival (Shakesby, 1997)
more than a few meters separating distinctly peri- rampart deposits. If the diamicton was transported
dotite from distinctly serpentinite diamicton. Differ- by glaciers, Bluff Lake ridges must be of Pleistocene
ences are readily recognizable from stones on the age, because Mohr et al. (2000) sampled a nearly
ground surface that have mostly rough yellowish continuous core from Bluff Lake on the proximal
brown coatings to the north and smooth light grey to side of the ridges from which they obtained a radio-
white coatings to the south. Mean densities of typi- carbon date of 11,550 years that has been calibrated
cal stone interiors were found to be 2.82 ± 0.10 Mg/ to about 13,500 years BP. On the other hand, the
m3 (n = 4) on the north and 2.56 ± 0.01 Mg/m3 (n = soils on the proximal ridge do not appear to be that
6 E. B. ALEXANDER

old, and the apparent draping of diamicton over ridges and the reddest colors are in the upper parts
older drift on the distal ridge is more characteristic of the cambic horizons. In serpentinite, representa-
of pronival rampart deposits. The older drift is tive sample colors are 10YR 5/3 dry (Rr = 3.0) and
assumed to be from Wisconsin glaciation, based on 10YR 4/3 moist (Rr = 3.8) on the late BL ridge and
descriptions by Sharp (1960) of Quaternary glacia- 9YR 5/3 dry (Rr = 3.6) and 9YR 4/3 moist (Rr = 4.5)
tion in the central part of the Klamath Mountains. on the early BL ridge. In peridotite, representative
Modes of diamicton transport and deposit, however, sample colors are 8YR 5/6 dry (Rr = 8.4) and 8YR
are not particularly important for the purposes of 4/5 moist (Rr = 8.8) on the late BL ridge and 7YR 5/
relating soil redness to parent materials, although it 6 dry (Rr = 9.6) and 7YR 4/5 moist (Rr = 10.0) on
would be interesting to learn more about rates of soil the early BL ridge.
reddening in peridotite and serpentinite materials. Soil redness data were compared within each
Soils on the proximal, or late, Bluff Lake ridge combination of ridge (early and late) and soil parent
are Entisols and Inceptisols in serpentinite diamic- material (peridotite and serpentinite) first. The only
ton and all are Inceptisols in peridotite diamicton. difference between means for lateral locations
Soils on the distal, or early, Bluff Lake ridge are all within these four ridge-PM combinations was that
Inceptisols, but with slightly redder and thicker the mean for the location nearest the parent material
cambic horizons than in comparable materials of the boundary was significantly greater (α < 0.05) than
late Bluff Lake ridge. Soils on both ends of the late for other locations in the serpentinite material on the
Bluff Lake ridge are assumed to be the same age, late Bluff Lake ridge. Mean redness ratings
but younger than soils on the early Bluff Lake ridge. appeared to be slightly lower on crest positions for
all ridge-PM combinations, but differences between
Methods means for slope positions were significant (α < 0.05)
Soils were sampled from 6 to 18 cm depths at 12, only in peridotite materials on the late Bluff Lake
24, and 36 meters laterally from the boundary sepa- ridge. There mean redness ratings were lower for the
rating peridotite and serpentinite diamictons, on crest position based on both dry and moist colors
both sides of the boundary. Lateral sampling was and higher for the lower slope position based on
repeated at crest, upper, middle, and lower slope moist colors only.
positions on both early and late Bluff Lake Differences between mean redness ratings for
moraines. Redness ratings were computed from soils with different parent materials were highly sig-
Munsell color designations as in the Rattlesnake nificant (α < 0.01) on both ridges, but differences
Creek terrane. Differences among means for sam- between means for early and late ridges were insig-
ples from different slope positions and lateral loca- nificant (α > 0.05) for both materials (Table 1). In
tions were compared by Fisher’s variance ratio (F) other words, the parent material effects are greater
and LSD (least significant difference). than the ridge, or age, effects. Only by lowering the
Composite samples consisting of equal amounts confidence level to 90% (α < 0.10) were the means
of fine earth from each of the three slope locations for all combinations of ridge and soil PM different.
within each combination of ridge, soil parent mate- Differences in Bluff Lake soil PM are indicated
rial (PM), and slope position were sieved to separate by concentrations of light and heavy magnetic sepa-
sand (0.062–2.0 mm), following bleach (Na- rates from fine sand (0.125 to 0.25 mm) fractions.
hypochlorite) and citrate-dithionite treatments to Mean concentrations (weight) of light separates are
remove readily oxidizable organic matter and free 36% in the peridotite and 63% in the serpentinite,
Fe oxides. Fine sand (0.125 to 0.25 mm) from the and those of heavy magnetic separates are 6% in the
samples was separated into heavy and light fractions peridotite and 13% in the serpentinite. Relation-
with bromoform (SG = 2.89). Magnetic grains were ships of soil redness ratings (moist) to the weight
separated with a hand magnet. Fine sand grains percentage of nonmagnetic heavy grains in fine sand
were identified and counted in oils of RI = 1.65 for (0.125–0.25 mm) fractions are shown in Figure 6.
the nonmagnetic heavy fraction and RI = 1.57 for Fine sand concentrations were statistically the same
the light fraction. for all four combinations of ridge and soil PM (α >
0.05), with a mean of 84 g/kg of fine earth. Mean
Results nonmagnetic heavy fractions within fine sand sepa-
The sampling depth was shallow, because cam- rates were different among the six comparisons of
bic horizons are thin in soils of the Bluff Lake (BL) the four combinations of ridge and soil PM (α <
 SERPENTINE SOIL REDNESS 7

TABLE 1. Redness Ratings of Soil Samples from 6 to 18 cm Depths at 48 Locations on the Bluff Lake (BL)
Ridges with Soil Parent Materials (PM) of Peridotite and Serpentinite Diamicton

Redness Rating1 Grouping of Classes2

Dry Moist α < 0.01 α < 0.05 α < 0.10

Early BL Ridge
Peridotite PM 7.5 8.8 a a a
Serpentinite PM 3.3 4.5 b b b

Late BL Ridge
Peridotite PM 6.5 7.5 a c c
Serpentinite PM 3.0 3.6 b b d

1Means of 12 samples in each class (4 slope position by 3 lateral locations).

2Class groupings are the same for ratings with both dry and moist soil samples.

FIG. 6. Heavy nonmagnetic fractions of fine sand (% by weight) and moist redness ratings for the four combinations
of two Bluff Lake ridges (early and late) and two soil parent materials (peridotite and serpentinite diamicton).

0.05), except that concentrations were somewhat from gabbro in the diamicton and from ash of Cas-
similar in peridotite diamicton on both ridges (0.05 cade Range volcanoes.
< α < 0.10). Redness ratings are obviously related to
nonmagnetic heavy mineral concentrations in fine Rationale for Serpentine Soil Redness
sand separates, even though relationships are con-
founded by different soil PM ages on the two ridges. Because peridotite and serpentinite have very
Minerals in the light fractions are mostly serpentine similar chemical compositions, weathering differ-
with minor feldspars, quartz, and glass, and those in ences must be related to differences in mineral com-
the heavy fraction, after removing magnetite, are position. And because soil redness is related to iron
mostly olivine and pyroxenes with minor horn- oxides (Scheinost and Schwertmann, 1999), it is
blende. The minor constituents are presumed to be particularly important to know in which minerals the
8 E. B. ALEXANDER

TABLE 2. Distribution of Iron in Minerals of Peridotite, or More Specifically Harzburgite,1 and in

Serpentinite with the Same Elemental Content (Serpentinite A) or after the Removal of Just Enough Mg
to Eliminate Brucite (Serpentinite B)2

Harzburgite Serpentinite A3 Serpentinite B4

Mineral Iron (Fe) in Mineral Iron (Fe) in Mineral Iron (Fe) in
content mineral rock content mineral rock content mineral rock

g/kg
Olivine 750 67 50.2
Enstatite 218 44 9.6
Diopside 26 14 0.4
Chromite 6 249 1.5 5 249 1.2 5 249 1.2
Lizardite 880 32 28.2 910 32 29.1
Chrysotile 50 10 0.5 52 10 0.5
Brucite 42 173 7.3
Magnetite 22 723 15.9 33 724 23.9

Sum 1000 61.7 999 53.1 1000 54.7

1From Page, 1967.

2Mass balance was obtained by the addition of 13%, or slightly more, water to form serpentinite.
3Chromite is assumed to be conserved during serpentinization, but its concentration is reduced because of the addition of

water in the process. The ratio of lizardite to chrysotile is on the order of 20 to 1 at Burro Mountain (Coleman and Keith,
1971). Chemical compositions of lizardite and chrysotile are from Coleman, 1971. The iron content of brucite is based on a
1:4 molar Fe:Mg ratio, which is in the range given by Page, 1967.
4A serpentinite composition based on just enough loss of Mg (1%, or 10 g/kg, about 4% of the Mg) to eliminate brucite.

iron is concentrated and to have some knowledge tite, and in many places brucite. Chromite was
about how readily these minerals weather to release assumed to persist during the hydration of peridotite
that iron. to serpentinite. Assuming that the serpentinite has
Peridotite in the Klamath Mountains is most the same chemistry as the harzburgite, except for
commonly harzburgite. Therefore, the mineralogical addition of water and loss of Ca during serpentiniza-
composition of a harzburgite is compared to a ser- tion, and utilizing lizardite and chrysotile analyses
pentinite with the same chemical composition, on an from Coleman (1971), proportions of lizardite,
anhydrous basis (Table 2). Although the harzburgite chrysotile, brucite, and magnetite were chosen to
is from Burro Mountain in the California Coast account for practically all of the Si and Mg and all of
Ranges, the chemistry and mineralogy are expected the Fe. Proportions of lizardite and chrysotile were
to be very similar in harzburgite of the Klamath chosen somewhat arbitrarily, because they both
Mountains. The harzburgite at Burro Mountain con- have nearly the same amounts of Si and Mg, but
tains olivine (forsterite), orthopyroxene (enstatite), Page (1967) indicated that the amount of lizardite is
clinopyroxene (diopside), and a spinel mineral that generally equal or greater than the amount of
is apparently picotite (Page, 1967). From the chem- chrysotile at Burro Mountain. He noted that Fe sub-
ical analysis of harzburgite in Coleman (1971), the stitutes for Mg in brucite and a 1:4 molar Fe:Mg
proportions of these minerals were estimated by ratio (173 g/kg Fe) was assumed in my computa-
allotting all Cr to chromite (rather than picotite), Ca tions. All remaining Fe, after its allocation to
to diopside, and the remaining Si, Mg, and Fe to oli- serpentine and brucite, was assigned to magnetite.
vine and enstatite. The main serpentinite minerals The results of mineral allocation in Table 2 show
at Burro Mountain are lizardite, chrysotile, magne- that most of the Fe in peridotite is in olivine. Olivine
 SERPENTINE SOIL REDNESS 9

FIG. 7. Redistribution of iron among minerals upon the serpentinization of three model peridotites, assuming the
addition of about 14% water during serpentinization. The olivine is assumed to be forsterite (Fo90) and the orthopyroxene
is assumed to be enstatite (En90). About 14% (140 g/kg) of the Si and all of the Ca from diopside must be removed in the
serpentinization of lherzolite to obtain mass balance.

is a relatively easily weathered mineral (Wilson, ite is in magnetite, which is very resistant to weath-
1975; Delvigne et al., 1979; Taylor and Eggleton, ering, and in serpentine minerals, which are
2001). It rapidly releases Fe, on a scale of 103 years, considerably more resistant to weathering than is
that can be oxidized to impart reddish colors to soils. olivine. Serpentinized dunite is an exception (Fig.
In contrast, most of the Fe in serpentinized peridot- 7). It lacks sufficient Si to use all of the Mg in the
10 E. B. ALEXANDER

formation of serpentinite. The excess Mg forms bru- Acknowledgments

cite, which is converted to pyroaurite or coalingite
I thank Rebecca Burt, Natural Resources Con-
in regolith or reacts with groundwater to release Fe
servation Service, Lincoln, California, and Ray-
and form hydromagnesite or artinite (Hostetler et al.,
mond (Bud) Burke, Humboldt State University,
1966; Mumpton and Thompson, 1975), none of Arcata, California, for very helpful reviews of the
which persist in soil environments. Brucite contains manuscript.
enough Fe that there is much less left for the forma-
tion of magnetite upon the serpentinization of dun-
ite. This is reflected in the relatively low magnetic REFERENCES
susceptibility of serpentinized dunite, compared to
serpentinized harzburgite (Toft et al., 1990). Alexander, E. B., 2000, A re-examination of serpentine
soil redness–bedrock specific gravity relationships:
Because the olivine in peridotite is considerably Soil Survey Horizons, v. 41, 24–26.
more weatherable than serpentine in most soil envi- ______, 2003, Trinity Serpentine Soil Survey: Mapping
ronments and much more weatherable than magne- transferred to a GIS: Unpubl. manuscript, USDA For-
tite, peridotite releases much more Fe more rapidly est Service, Shasta-Trinity National Forest, Redding,
than does serpentinite. The weathering of brucite in California.
serpentinized dunite releases Fe rapidly, but about Alexander, E. B., Adamson, C., Graham, R. C., and Zinke,
P. J., 1990, Mineralogy and classification of soils on
half of the bulk-rock Fe is still retained in the ser- serpentinized peridotite of the Trinity ophiolite, Cali-
pentine. A few shallow and moderately deep soils on fornia: Soil Science, v. 149, 138–143.
serpentinite have quite reddish colors, at least in Alexander, E. B., Adamson, C., Zinke, P. J., and Graham,
5YR hues. Could their serpentinite parent materials R. C., 1989, Soils and conifer forest productivity on
have had high concentrations of brucite? No one has serpentinized peridotite of the Trinity ophiolite, Cali-
investigated this possibility. fornia: Soil Science, v. 148, 412–423.
Coleman, R. G., 1971, Petrologic and geophysical nature
Not only must weathering release Fe to soils for of serpentinites: Geological Society of America, Bulle-
them to redden, but the Fe must be oxidixed to form tin, v. 82, p. 987–918.
ferrihydrite (about 6 to 7YR hues) or minerals such Coleman, R. G., and Keith, T. E., 1971, A chemical study
as goethite (about 10YR hues) and hematite (about of serpentinization—Burro Mountain, California: Jour-
1 to 2YR hues, 5R to 5YR range). Well drained soils nal of Petrology, v. 12, part 2, p. 311–328.
Delvigne, J., Bisdom, E. B. A. , Sleeman, J., and Stoops,
generally contain more than one of these minerals,
G., 1979, Olivines, their pseudomorphs and secondary
allowing a broad range of colors from grey to red products: Pedologie, v. 29, 247–309.
(Schwertmann, 1993; Scheinost and Schwertmann, Hostetler, P. B., Coleman, R. G., Mumpton, F. A., and
1999). Evans, B. W. , 1966, Brucite in alpine peridotites:
Soil redness differences have been attributed to American Mineralogist, v. 51, 75–98.
Irwin, W. P., 1994, Geologic map of the Klamath Moun-
soil temperature and moisture conditions (Schwert-
tains, California and Oregon: U.S. Geological Survey,
mann, 1993), as well as to soil parent material dif- Miscellaneous Investigations Series, Map I-2148.
ferences. There is no evidence of any temperature or Mohr, J. A., Whitlock, C., and Skinner, C. N., 2000, Post-
moisture effects in the Rattlesnake Creek terrane, glacial vegetation and fire history, eastern Klamath
where these are related to altitude. Neither dry nor Mountains, California: The Holocene, v. 10, p. 587–
moist redness ratings of peridotite (n = 50) or ser- 601.
Mumpton, F. A., and Thompson, C. S., 1975, Mineralogy
pentinite (n = 76) soils were correlated with altitude
and origin of the Coalinga asbestos deposit: Clays and
in the range from 650 to 1760 meters above sea level Clay Minerals, v. 23, p. 131–143.
r < 0.3, r2 < 0.1). Even the relatively young soils at Page, N. J., 1967, Serpentinization at Burro Mountain,
about 1930 m altitude on the Bluff Lake ridges have California: Contributions to Mineralogy and Petrology,
relatively red colors—at least in peridotite materi- v. 14, p. 321–342.
als, which weather much more rapidly than the ser- Scheinost, A. C., and Schwertmann, U., 1999, Color iden-
tification of iron oxides and hydroxysulfates: Use and
pentinite materials. Soil parent material is clearly
limitations: Soil Science Society of America, v. 63, p.
the dominant factor in the reddening of serpentine 1463–1471.
soils in the Klamath Mountains, with the exceptions Schwertmann, U., 1993, Relations between iron oxides,
of very young and very old soils. soil color, and soil formation: in Bigham, J. M., and
 SERPENTINE SOIL REDNESS 11

Ciolkosz, E. J., eds., Soil color: Soil Science Society of ______, 1999, Soil taxonomy: A basic system of soil clas-
America, Special Publication 31, 51–69 . sification for making and interpreting soil surveys:
Shakesby, R. A., 1997, Pronival (protalus) ramparts: A Washington, DC, U.S. Department of Agriculture, Soil
review of forms, processes, diagnostic criteria, and Conservation Service, 869 p.
paleoenvironmental implications: Progress in Physical Taylor, G., and Eggleton, R.A., 2001, Regolith geology and
Geography, v. 21, 394–418. Geomorphology: New York, NY, Wiley.
Toft, P. B., Arkani-Hamed, J., and Haggerty, S. E., 1990,
Sharp, R. P., 1960, Pleistocene glaciation in the Trinity The effects of serpentinization on density and mag-
Alps of northern California: American Journal of Sci- netic susceptibility; A petrographic model: Physics of
ence, v. 258, 305–340. the Earth and Planetary Interiors, v. 65, p. 137–157.
Soil Survey Staff, 1993, Soil survey manual: Washington, Wilson, M. J., 1975, Chemical weathering of some primary
DC, U.S. Department of Agriculture, Handbook No. rock-forming minerals: Soil Science, v. 119, p. 349–
18. 355.

View publication stats