About the Cover

The cover: The inset foreground photograph is a serpentinite rock covered with intensely iron oxide pigmented clays. The clays have translocated into the rocks along shear planes. The broken edge at the upper end of the rock displays the rocks' interior color which is considerably different than the clay and iron oxide produced from these rocks. The background picture is of a serpentinite rock from the surface of serpentinite soil displaying warty surface texture due to differential weathering of the minerals that make up the rock. The rock is bisected diagonally by a chrysotile vein. Please see pages 2087–2095, “Plant-Available Calcium Varies Widely in Soils on Serpentinite Landscapes” by Donald G. McGahan, Randal J. Southard, and Victor P. Claassen

SSSAJ: Volume 73: Number 6 • November–December 2009 2087

This research was supported in part by Hatch project CA-D*-LAW-4525-H, California Department of Transportation, The David and Lucile Packard Foundation Interdisciplinary Science Program, and The Andrew W. Mellon Foundation Conservation and Environment Program.

Soil Sci. Soc. Am. J. 73:2087-2095doi:10.2136/sssaj2008.0087Received 10 Mar. 2008.*Corresponding author (mcgahan@tarleton.edu).© Soil Science Society of America677 S. Segoe Rd. Madison WI 53711 USAAll rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Ultramafi c rocks, of which serpentinite is a member, occupy <1% of the land surface of the earth, but they have long been

the study of botanists, plant physiologists, phytochemists, plant geographers, and other earth and life scientists. Serpentinite-derived soils oft en support vegetation that stands in stark contrast (reduced vigor, reduced numbers, and more endemic species) to vegetation on adjacent non-serpentinite-derived soils. Th e per-ceived diff erence has oft en been termed the serpentine factor or syn-drome, and botanists oft en refer to “serpentine fl ora” (Alexander et al., 2007; Brooks, 1987; Jenny, 1980; Kruckeberg, 1985).

Th e term serpentine soil, as commonly used in the literature, in many cases is a misnomer. Serpentine is the name of a class of minerals of which antigorite, chrysotile, and lizardite are

the most recognized members (Deer et al., 1997). Serpentines are 1:1 minerals with the general formula Mg3Si2O5(OH)4 and are trioctahedral analogs of the dioctahedral kaolinite [Al2Si2O5(OH)4]. Serpentine minerals are not formed in soils. Th ey are the products of metamorphism (serpentinization) of ultramafi c rocks such as dunites, pyroxenites, and peridotites (O’Hanley, 1996). Ultramafi c rocks with abundant serpentine are defi ned as serpentinite. Th ese serpentinite rocks usually also contain minor amounts of the minerals magnetite, brucite, talc, and various carbonates. Many other accessory minerals may occur in serpentinites as trace components, including pyroxenes, am-phiboles, garnets, chlorite, vermiculite, chromite, ilmenite, and titanite. A preferable term to serpentine soil might therefore be serpentinite-derived soil to distinguish between the mineral and rock names.

Th e perceived diff erences between soils on what we nomi-nally call serpentinite and soils adjacent to them, and perceived to be non-serpentinite derived, have perhaps led to a perception that the soils formed on serpentinite landscapes are relatively ho-mogenous with respect to the “serpentine factor.” Th is perception masks inherent variability in serpentinites, in the soils formed from those serpentinites, and in the plants that grow in the soils.

Serpentinites are metamorphic rocks produced by metaso-matic metamorphism. Metasomatism is a process of indefi nite elemental replacement, loss, or addition of elements as a result of percolating solutions (Merrill, 1906). As a result, during meta-somatic metamorphism, the anhydrous minerals of the protolith dunite, pyroxenite, or peridotite become more hydrous, and Ca content decreases from the original rocks, resulting in a relative enrichment of Mg in the serpentinite and, generally, in soils

Donald G. McGahan*Dep. of Agribusiness, Agronomy, Horticulture and Range ManagementTarleton State Univ. and AgriLIFE Research and Extension CenterStephenville, TX 76402

Randal J. SouthardVictor P. ClaassenDep. of Land, Air and Water ResourcesUniv. of CaliforniaDavis, CA 95616

FOR

EST, RA

NG

E & W

ILDLA

ND

SOILS

Plant-Available Calcium Varies Widely in Soils on Serpentinite Landscapes

Serpentinitic landscapes are characterized by unusual plant species composition and vegetation patchiness. We hypothesized that soil extractable Ca content is correlated with vegetation Ca concentration and may contribute to the unusual vegetation characteristics. Eleven pedons from California were sampled within soil survey mapping polygons containing the modal location for a classic “serpentine soil.” Soil NH4OAc-extractable Ca/Mg ratio varied from 0.1 to 1.5 and total elemental Ca/Mg ratio varied from <0.01 to 1.9. Soil extractable Ca content was infl uenced by parent material mineralogy: soils with a Ca/Mg ratio of 0.2 or less were derived from serpentinite parent materials with only trace Ca-bearing minerals; soils with a Ca/Mg ratio >1.0 were derived from non-serpentinite parent materials, and soils with a Ca/Mg ratio from 0.2 to 1.0 were derived from serpentinite parent material that had minor amounts of accessory Ca-bearing minerals. Based on greenhouse studies, the Ca concentration of the leaf of the grass Vulpia microstachys (Nutt.) Monro correlated better with soil-extractable Ca (r2 = 0.69, P < 0.01) than with total elemental analysis Ca (r2 = 0.40, P > 0.05). Th e shallow soil depths of the pedons, together with the generally high coarse fragment content, limit the soil volume available to roots. Pedons with the highest extractable Ca/Mg (1.34 and 1.89) also had the greatest extractable Ca pool, 1169 and 394 g m−2, respectively. Among the 11 pedons, extractable Ca/Mg ratio, extractable-Ca pool, and the distribution of the Ca pool within the pedon all varied, and these variables were important determinants of plant performance, as determined by leaf Ca content. Of these variables, soil extractable-Ca content was the best determinant of plant performance as indicated by leaf Ca concentration.

Abbreviations: CEC, cation exchange capacity; XRD, x-ray diff raction.

2088 SSSAJ: Volume 73: Number 6 • November–December 2009

formed from the serpentinite (Coleman and Keith, 1971; Page, 1966, 1967). Th e Ca/Mg ratio has been used to characterize the high-Mg and low-Ca condition that exists in these soils. Early researchers attributed poor growth to Mg toxicity on serpenti-nite-derived soils (Loew and May, 1901; Gordon and Lipman, 1926); however, Vlamis and Jenny (1948) demonstrated that Ca defi ciency, rather than Mg toxicity, was the primary cause of poor plant growth. Walker et al. (1955) emphasized the importance of Ca to vegetation and demonstrated that non-native species yields were reduced about 50% when exchangeable Ca was 20% or less of the cation exchange capacity (CEC), and the plants had little or no growth or yield below 10% exchangeable Ca. Native plant species, however, were better able to extract Ca, and yield was only reduced 24% in the 5 to 3% exchangeable Ca range.

Many soils formed on “serpentinite landscapes” suff er from low Ca content and an abundance of Mg, but this is clearly not always the case. McGahan et al. (2008) described serpentinite parent materials that had trace accessory Ca-bearing minerals that infl uenced the surface extractable Ca/Mg ratio, and other parent materials in the “serpentinite landscapes” that were not serpentinites. Soil or geologic maps depicting serpentinite land-scapes oft en do not capture the inherent variability of rock and soil properties in these landscapes.

A Ca/Mg ratio of 0.7 is oft en adopted as a quotient to sepa-rate serpentinite parent material from non-serpentinite parent material (Alexander et al., 1985; Brooks, 1987). In light of the work by Walker et al. (1955) and Vlamis and Jenny (1948), this value may not be particularly useful when working with “native” serpentinite vegetation. Proctor and Woodell (1971) found a wide range in the Ca/Mg ratio (0.03–2.37) in “serpentine” soils from Scotland and England. Rabenhorst and Foss (1981), in an eff ort to predict mafi c or ultramafi c parent lithology while map-ping soils of the eastern Piedmont of Maryland, found in a popu-lation of 39 pedons, that if the exchangeable Ca/Mg ratio was < 0.1 the probability that the soil was formed from serpentinite rather than a mafi c lithology was 98%, a Ca/Mg ratio of 0.2 to 0.3 yielded a serpentinite probability of 53%, and at 0.6 to 0.7 Ca/Mg ratio, the probability was only 21%. Clearly, at extract-able Ca/Mg ratios >0.2 the confi dence that the soil is derived from ultramafi c parent material (e.g., serpentinite) is low.

Th e use and understanding of the Ca/Mg ratio in serpen-tinitic landscapes is complicated by literature with references to Ca/Mg ratios without stating the method of determining the elemental concentrations, e.g., whole rock elemental analy-sis (Capm/Mgpm) vs. total soil fi ne earth fraction elemental analysis (Cast/Mgst) vs. soil extractable Ca and Mg (Cae/Mge). Investigators of the “serpentine eff ect” were warned by Proctor and Woodell (1971) not to rely on whole rock analysis as rep-resentative of the pool of soil nutrients available to plants. A link back to the parent material is probable, but Alexander et al. (2007) pointed out that geologic variation infl uencing veg-etation variation has not been examined in detail. While soil ex-tractable Ca and Mg contents are expected to give a better repre-sentation of the pool of soil nutrients available to plants than are total elemental analyses, additional characterization of the soil and rock material, including parent material elemental analyses, fi ne earth fraction elemental analysis, polarized light microscopy, x-ray diff raction (XRD), and energy dispersive x-ray spectroscopy,

are useful for identifying and characterizing serpentinite-derived soils (McGahan et al., 2008).

Even when careful characterization of the underlying par-ent material is performed, the determination of soil and parent material on the “serpentine eff ect” can be confounded by the occurrence of non-serpentinite parent materials on the serpen-tinitic landscape (Page, 1966, 1967; Coleman and Keith, 1971; McGahan et al., 2008). Th is is because upslope tectonic inclu-sions of non-serpentine rocks could infl uence Ca fertility via Ca-enriched pore waters from upslope weathering of Ca-containing minerals, or from colluvium containing Ca-bearing minerals. Eolian infl uences are also a possibility not fully explored by re-searchers investigating the “serpentine eff ect.”

We investigated soils formed on serpentinitic landscapes to test our hypothesis that these soils, nominally called serpentine soils, have a range of morphologic, chemical, and parent mate-rial properties, and that this variation is related to the Cae/Mge ratio, soil extractable Ca (Cae), and total soil Ca (Cast) contents. We present both Cast and soil total Mg (Mgst) and Cae and soil extractable Mg (Mge) from 11 soils, an admittedly limited data set, to convey the concept of Ca pools. Th e infl uence of diff er-ing amounts of Ca-bearing minerals is based on previous work (McGahan et al., 2008) where we recognized soils derived from (i) serpentinite with no more than trace amounts of Ca-bearing minerals as detected by XRD, (ii) non-serpentinite, and (iii) serpentinite with minor amounts of Ca-bearing minerals as de-tected by XRD. We then present Cast, Mgst, Cae, Mge, the Cae/Mge ratio, and the Cast/Mgst ratio vs. plant leaf Ca (Cap) content to test the hypothesis that the Cae content is correlated to the Cap concentration.

MATERIALS AND METHODSField

We used soils mapped as the Henneke soil series (clayey-skeletal, magnesic, thermic Lithic Argixerolls) and soil survey modal locations as a basis for sampling locations (Soil Survey Staff , 1999). Henneke soils are widely mapped, and are “soils formed in material weathered from serpentine and rocks of similar mineralogy” (Soil Survey Staff , 2007). Th e sampling sites encompassed a wide range of serpentinite landscape lithology. Th e parent materials of the soils used in this study have been examined for mineralogic composition in other studies so that mean-ingful relationships can be drawn from morphologic observations (McGahan, 2007; McGahan et al., 2008). All pedons were sampled on summits above modal locations to eliminate colluvial and alluvial infl u-ences, and were within the mapping polygon containing the Henneke series modal location for each soil survey area. Eleven sites were sampled across California (Fig. 1). Th e use of Henneke survey modal locations generally constrained the soil climate to be xeric (soil moisture regime) and thermic or mesic (soil temperature regime). Pits were excavated by hand tools. Th e soils were described and sampled using conventional procedures (Soil Survey Division Staff , 1993). Th e volume of cobbles and stones was estimated in the fi eld. Th e fi ne earth fraction and gravels were sampled in conjunction with compliant cavity bulk density mea-surements (Soil Survey Staff , 2004).

LaboratorySoils were air dried, sieved to pass through a 2-mm sieve, and the

fi ne earth fraction was analyzed for particle size distribution by the pipette method as described by Gee and Bauder (1986). Fine earth

SSSAJ: Volume 73: Number 6 • November –December 2009 2089

fraction (soil) extractable cations were measured by displacement with pH 7.0, 1 mol L−1 NH4OAc, and the concentrations of Na, Ca, Mg, and K in the leachate were measured with fl ame atomic absorption or emission spectrometry. Th e CEC was measured by saturating samples with 1 mol L−1 NH4OAc, at pH 7, washing with 95% ethanol, extract-ing with 2 mol L−1 KCl to remove adsorbed NH4 (Soil Survey Staff , 2004), and determining the NH4 concentration in the leachate conduc-timetrically (Carlson, 1978). Th e pH of the soil was determined at a 1:1 soil/water content on a mass basis (Soil Survey Staff , 2004).

Th e following materials (approximately 10-g subsamples of each) were analyzed for Cast and Mgst: the soil material from each horizon, the blended soil that was used for the greenhouse experiment (see below), and rock from the R or Cr horizon below each soil that was ground and sieved until all passed a 100-μm (no. 140) sieve. Elemental analysis of Ca and Mg was performed by inductively coupled plasma-atomic emission spectrometry, following a LiBO2 fusion, and a dilute HNO3 digestion (Sawhney and Stilwell, 1994).

Th e volume of gravels was determined directly by volume displace-ment in water. Displacement was performed aft er sieving compliant cav-ity bulk density samples to separate the gravels, then soaking the gravels for 15 h in deionized water, washing to remove slaked soil material, and drying at 105°C for 15 h (Soil Survey Staff , 2004).

GreenhousePlant roots were observed in the A and B horizons during fi eld

investigations, and we assumed that the roots integrate the chemical environment throughout the rooting zone. Th erefore, we blended soil in proportion to the soil of the A and B horizons for each pedon in our greenhouse pot studies. Th ese blended solum soil materials were used to fi ll 500-cm3 pots and planted with approximately three dozen seeds of Vulpia microstachys (Nutt.) Munro per pot. Vulpia microstachys was observed to grow at many serpentinite sites surveyed, and the seed used was collected from serpentinite landscapes. Aft er 3 wk of growth, the plants were fertilized with 0.5 L of solution equivalent to ¼ Hoagland solution of the elements N and P (4000 and 500 μmol L−1, respectively) each week to remove potential eff ects of defi ciencies of these elements. Aft er 16 wk of growth and before the plants set seed, the aboveground plant material, hereaft er referred to as leaf, produced in each pot was clipped, rinsed in deionized water, dried at 45°C for 72 h, ground, weighed, and ashed by heating in a muffl e furnace at 500°C for 4 h; the ash was digested in hot 1 mol L−1 HCl, and the digest analyzed for Mgp and Cap by atomic absorption spectrometry (Gavlak et al., 1994). Leaf Cap content was used as the indicator of plant performance.

Regression relationships between soil values (Cast, Cae, Mgst, and Mge) and plant leaf values were determined by minimizing the sum of the squared dif-ferences between the observed and predicted values using the Marquardt–Levenberg algorithm, and all statistical analyses were performed using SigmaPlot for Windows, Version 8.02 (SPSS Inc., Chicago, IL).

RESULTS AND DISCUSSIONPedon Morphologies

Soil classifi cation (Table 1) refl ected the variation in soil morphology, chemistry, and mineralogy among the 11 pedons (Table 2), even when the soil sampling sites were con-strained to polygons containing the modal loca-tion of the same soil series (Henneke).

All the pedons sampled were lithic at the subgroup level or were in shallow families. Textures were generally loamy (loams, sandy loams, sandy clay loams, and clay loams), with the excep-tion of clay textures at the base of the Colusa County pedon (Bt2 horizon) and in the B horizons of the Shasta-Trinity National Forest pedon (Table 2). Four pedons, Shasta-Trinity National Forest and Napa, Colusa, and Tehama counties, were not skeletal (Table 1). Th ese textural deviations resulted in loamy or clayey family textural classes. Other variations in soil properties that re-sulted in alternate classifi cation of the soils were the presence of an ochric epipedon (rather than mollic) in the Napa, Shasta, and San Benito counties sites and the absence of an argillic horizon in the Napa and Shasta counties and Los Padres National Forest pedons. Th erefore, the San Benito County site was classifi ed as an Alfi sol, the Napa and Shasta County sites as Inceptisols, and the Los Padres National Forest site as a Haploxeroll (Soil Survey Staff , 1999).

Th e shallow soils and high coarse fraction percentage of-fer a limited soil volume from which vegetative communities

Fig. 1. Location of soil sampling sites on ultramafi c parent materials (serpentinites) and soil survey areas in California.

Table 1. Soil site classifi cation determined from site pedon morphology.

Soil survey Taxonomic classifi cation

Colusa County clayey, magnesic, thermic, shallow Typic ArgixerollGlenn County loamy-skeletal, magnesic, mesic Lithic Argixeroll†

Kings County loamy-skeletal, magnesic, mesic Lithic Argixeroll†

Napa County loamy, mixed, active, thermic Lithic Haploxerept

San Benito County loamy-skeletal, magnesic, mesic, shallow Typic Haploxeralf†

Shasta County loamy-skeletal, magnesic, thermic Lithic Haploxerept

Tehama County loamy, mixed, superactive, mesic, shallow Typic Argixeroll†

Mariposa County loamy-skeletal, magnesic, thermic Lithic Argixeroll

Mendocino County loamy-skeletal, magnesic, thermic Lithic Argixeroll

Shasta-Trinity National Forest clayey, magnesic, mesic, shallow Typic Argixeroll†Los Padres National Forest loamy-skeletal, magnesic, mesic Lithic Haploxeroll†† Elevation of the site (about 1000 m) suggested that the soil temperature regime was mesic but no seasonal temperature measurements were made. Soils were mapped as thermic soil temperature regime.

2090 SSSAJ: Volume 73: Number 6 • November–December 2009

can assimilate nutrients. Since it is well established that low Ca availability is a primary limitation to vegetative vigor (Walker et al., 1955), we focused on the distribution of Ca in the extractable fraction, Cae, in relation to the total pool of these elements.

Our sampling strategy excluded sites that could be infl u-enced by colluvium, alluvium, or from Ca dissolved in pore water as a result of upslope weathering of Ca-enriched parent materials. Diff erences in the Cae/Mge ratio were attributed to the diff er-ences in the parent material, Capm and Mgpm, or to the relative

weathering of the Mg- and Ca-bearing minerals that comprised the parent material.

For seven of the 11 pedons, Mg was the dominant extract-able cation, and those soils had a Cae/Mge ratio of <0.2 (Table 3). Four pedons, from Napa, Tehama, Kings, and Shasta counties, were in contrast to the dominance of Mge and had a Cae/Mge ra-tio of >0.2 in most horizons of the pedon, which was consistent with our hypothesis that soils formed on serpentinitic landscapes would have variability in the Cae/Mge ratio (Table 3).

Table 2. Extractable cations, cation exchange capacity (CEC), and particle size distribution of soils from serpentinite-derived landscapes.†

Horizon Depth K‡ Mg‡ Ca‡ CEC-7§ Cae/Mge‡ Sand Silt Clay

cm —–––––––––––––—– cmol kg−1 ——––––––––––-— –––––––––——– % ——––––––––––Colusa County

A 0–3 0.3 25.7 9.0 23.0 0.35 46 33 21

ABt 3–8 0.2 23.7 5.6 21.4 0.23 42 31 27

Bt1 8–30 0.1 23.2 1.9 23.7 0.08 40 28 32

Bt2 30–42 0.1 44.3 0.3 43.0 0.01 27 18 55

Glenn County

A 0–9 0.2 21.9 4.4 18.8 0.20 54 20 25

Bt 9–34 0.3 18.9 4.5 27.5 0.20 49 18 34

Kings County

A 0–4 0.2 12.1 1.4 23.7 0.58 43 55 3

Bt1 4–13 0.6 12.1 7.0 18.0 0.51 36 49 15

Bt2 13–40 0.8 17.7 4.4 20.0 0.25 27 45 27

Napa County

A 0–6 0.4 16.4 24.1 31.1 1.47 43 42 15

Bt 6–20 0.1 16.2 21.5 27.9 1.32 54 29 17

San Benito County

A 0–8 < 0.1 3.0 0.7 3.1 0.24 80 8 12

Bt 8–20 0.0 0.1 5.6 5.6 0.11 74 7 19

Shasta County

A 0–3 0.3 9.7 3.5 13.2 0.36 37 49 14

Bt 3–40 0.2 10.1 2.0 12.2 0.20 39 46 14

Tehama County

A 0–3 0.4 12.4 16.3 40.2 1.31 57 19 24

Bt1 3–7 0.5 20.2 16.5 32.7 0.82 55 20 25

Bt2 7–16 0.2 27.7 13.7 30.0 0.50 51 17 32

Mariposa County

A 0–3 0.6 28.3 5.8 29.3 0.20 42 26 32

Bt1 3–14 0.4 33.7 6.2 27.9 0.18 41 34 25

Bt2 14–34 0.3 44.3 6.2 34.1 0.14 36 33 31

Bt3 34–49 0.1 78.2 4.4 43.6 0.06 36 31 33

Mendocino County

A 0–2 0.5 12.8 2.2 19.3 0.18 57 30 13

Bt1 2–8 0.2 27.1 3.4 19.8 0.13 51 28 21

Bt2 8–30 0.2 30.5 3.7 22.9 0.12 51 26 23

Shasta-Trinity National Forest

A1 0–2 0.2 18.8 6.6 23.0 0.35 34 44 22

A2 2–5 0.2 25.5 4.1 21.5 0.16 33 42 25

AB 5–12 0.3 29.3 3.5 29.6 0.12 24 36 40

Bt1 12–23 0.2 30.3 1.5 26.5 0.05 20 29 51

Bt2 23–40 0.1 64.9 0.8 48.7 0.01 11 22 67

Los Padres National ForestA 0–18 0.2 12.1 1.4 10.1 0.11 64 12 24† Na was measured but was 0.1 cmol kg−1 or less in all cases except the San Benito County Bt horizon (0.6 cmol kg−1). ‡ Extracted by 1 mol L−1 NH4OAc at pH 7; Cae = extractable Ca; Mge = extractable Mg.§ The CEC-7 may underestimate the actual CEC as both the index cation (NH4

+) and the displacing cation (K+) can cause collapse of vermiculite, and vermiculite was present in most horizons.

SSSAJ: Volume 73: Number 6 • November –December 2009 2091

Table 3. Soil bulk density (Db), coarse fraction volume (CF), extractable Ca displaced with 1 mol L−1 NH4OAc at pH 7 (Cae), and fi ne earth fraction (<2 mm) total Ca (Cast) and Mg (Mgst) for each survey area.

Horizon Db† CF‡ Cae Cae Cast Mgst Cae/Cast Cast/Mgst

g cm−3 cmol m−2 ———– g m−2 ———Colusa County

A 1.05 ± 0.08 (3) 0.12 249 50 610 2559 0.08 0.24

ABt 1.21 ± 0.02 (3) 0.27 247 50 833 4493 0.06 0.19

Bt1 1.33 ± 0.13 (3) 0.27 406 81 3740 22504 0.02 0.17

Bt2 1.05 ± 0.19 (2) 0.16 32 6 1755 10264 < 0.01 0.17

Total 934 187 6939 39820 0.03 0.17

Glenn County

A 1.09 ± 0.12 (3) 0.55 194 39 142 6669 0.27 0.02

Bt 1.14 ± 0.37 (3) 0.84 205 41 143 6454 0.29 0.02

Total 399 80 285 13123 0.28 0.02

Kings County

A 0.80 ± 0.17 (3) 0.50 22 4 279 556 0.02 0.50

Bt1 1.15 ± 0.04 (3) 0.75 181 36 457 910 0.08 0.50

Bt2 1.06 ± 0.05 (3) 0.67 595 119 1444 3417 0.08 0.42

Total 799 160 2180 4883 0.07 0.45

Napa County

A 1.51 ± 0.29 (3) 0.02 2140 429 2659 1879 0.16 1.41

Bt 1.32 ± 0.35 (4) 0.07 3695 740 5122 3907 0.15 1.31

Total 5835 1169 7781 5787 0.15 1.34

San Benito County

A 1.51 ± 0.32 (3) 0.43 255 51 340 14650 0.15 0.02

Bt 1.16 ± 0.20 (3) 0.60 39 8 306 11769 0.03 0.03

Total 294 59 646 26419 0.09 0.02

Shasta County

A 0.53 ± 0.33 (2) 0.78 48 10 53 393 0.18 0.14

Bt 0.94 ± 0.11 (2) 0.88 146 29 692 4749 0.04 0.15

Total 194 39 746 5143 0.05 0.14

Tehama County

A 0.83 ± 0.04 (3) 0.43 58 12 1216 858 0.01 1.42

Bt1 1.18 ± 0.06 (3) 0.10 692 139 3862 2413 0.04 1.60

Bt2 1.24 ± 0.16 (2) 0.34 1215 244 7707 3478 0.03 2.22

Total 1966 394 12785 6750 0.03 1.89

Mariposa County

A 1.07 ± 0.07 (3) 0.25 140 28 236 2793 0.12 0.08

Bt1 1.43 ± 0.05 (3) 0.38 605 121 990 12204 0.12 0.08

Bt2 1.63 ± 0.25 (2) 0.71 586 117 1061 10929 0.11 0.10

Bt3 1.51 ± 0.06 (2) 0.67 329 66 2169 5423 0.03 0.40

Total 1659 333 4455 31350 0.08 0.14

Mendocino County

A 0.98 ± 0.15 (3) 0.16 33 7 202 2517 0.03 0.08

Bt1 1.46 ± 0.08 (3) 0.14 166 33 635 12753 0.05 0.05

Bt2 1.24 ± 0.46 (2) 0.80 186 37 456 9022 0.08 0.05

Total 384 77 1294 24292 0.06 0.05

Shasta-Trinity National Forest

A1 0.93 ± 0.21 (3) 0.17 102 20 141 1539 0.15 0.09

A2 1.12 ± 0.13 (3) 0.13 120 24 280 2743 0.09 0.10

AB 1.06 ± 0.11 (3) 0.09 236 47 434 7159 0.11 0.06

Bt1 1.01 ± 0.01 (3) 0.05 158 32 392 11973 0.08 0.03

Bt2 1.05 ± 0.09 (3) 0.05 136 27 776 12354 0.04 0.06

Total 752 151 2023 35767 0.07 0.06

Los Padres National Forest

A 1.18 ± 0.13 (3) 0.46 69 14 262 19132 0.05 0.01Total 69 14 262 19132 0.05 0.01† Mean ± standard deviation, with number of samples in parentheses.‡ Volume fraction of horizon volume.

2092 SSSAJ: Volume 73: Number 6 • November–December 2009

Parent Material Infl uences on the Calcium/Magnesium Ratio

We examined the parent material from the R or Cr layer from each site by total elemental analysis, polarizing light mi-croscopy, back scatter electron microscopy, energy dispersive electron microscopy, and x-ray diff raction to document the vari-ability of the lithology. More detailed descriptions are reported elsewhere (McGahan, 2007; McGahan et al., 2008). Parent ma-terial mineralogic composition, or rock type, clearly aff ected the Cae/Mge ratio (McGahan et al., 2008).

Nine of the 11 parent materials were serpentinites. Th e Napa and Tehama county parent materials were quite diff erent from each other and from the nine other sites. Neither parent rock con-tained serpentine minerals and were therefore not serpentinites. Th e Napa County parent material contained dominantly vermicu-lite and albite, with minor amounts of Ca-bearing clino-pyroxene. Th e Tehama County parent material was dominated by grossular-ite, a calcsilicate ugrandite garnet, with subdominant amounts of the Ca-bearing sorosilicate, pumpellyite, and Ca-bearing clinopy-roxene. Th e rocks from the Shasta and Kings county sites were serpentinite, dominated by serpentine minerals, but had minor amounts of Ca-bearing accessory minerals (calcic clinoamphibole [tremolite] and calcsilicate ugrandite garnet [andradite]). Th e sev-en other parent materials were serpentinites and exhibited, at most, trace amounts of Ca-bearing minerals.

Extractable Calcium PoolsTh e Cae/Mge ratio (Table 2) broadly refl ected parent mate-

rial composition (McGahan, 2007; McGahan et al., 2008). Th e Cae/Mge values for Napa County were the greatest (1.3–1.5), followed by Tehama County (0.5–1.3). Shasta (0.3–0.4) and Kings (0.3–0.6) counties Cae/Mge values were intermediate, and the remaining sites generally had Cae/Mge values ≤0.2.

Th e total Cae pool, calculated from the Cae of the fi ne earth fraction of each horizon, in Napa County was nearly three times greater (1169 g m−2) than the next largest Cae pool, which was in Tehama County (394 g m−2) (Table 3). Th e variations in the Cae/Mge ratio were further reinforced by an increased Cae pool size where the Cae/Mge is >1 in Napa and Tehama counties. Th e fraction of Cae as a proportion of the Cast was also nearly fi ve times greater in the Napa County pedon (0.15) than in the

Tehama County pedon (0.03) (Table 3). Th is probably refl ects the relative weathering of the minerals present in the parent material. Th e Tehama County parent material contained Ca-bearing garnets, and garnets have greater resistance to weathering than the Ca-bearing clino-pyroxene of the Napa County parent material (Pettijohn, 1941).

We noted that the Shasta and Kings counties parent mate-rials had minor inclusions of Ca-bearing minerals. Th e Shasta County Cast/Mgst ratio (Table 3) was lower than the Cae/Mge ratio (Table 2). Quideau et al. (1999) used mass balance to identify Ca biocycling within natural chaparral ecosystems in California. We interpret the higher Cae/Mge ratio in the Shasta County pedon to be a refl ection of the biocycling of Ca.

Kings County had a greater proportion of Cast (0.25–0.50) than Shasta County (0.12–0.22) (Table 3). Th e Kings County Cast/Mgst ratio (Table 4) is similar to the Cae/Mge ratio (Table 3). Not only does the plant have a greater volume of fi ne earth fraction to draw from, but the CEC is greater in the Kings County soil (18–23 cmol kg−1) than the Shasta County soil (12–13 cmol kg−1). Th is suggests that the larger Cae pool asso-ciated with cation exchange on the mineral phase in the Kings County pedon may decrease the need for organic matter Ca as-sociations and a decreased intensity of Ca biocycling.

Th e Mariposa County site stands out among the other pe-dons because the Cae pools were also large, but the larger Cae pool was accompanied by a larger Mge pool, resulting in a low Cae/Mge ratio and a low Cast/Mgst ratio, similar to the other pedons.

Th ese pools of Ca probably vary as a function of parent ma-terial composition diff erences, mineral weathering diff erences, and biocycling. A number of processes can be inferred by inspec-tion of the Cast and Cae pools. If the Cae fraction of the Cast increases toward the surface, this may refl ect biocycling via litter deposition and decomposition, and Ca retention in surface ho-rizons. If the ratio increases with depth, this may indicate Ca-bearing mineral weathering and less intensive biocycling of Ca in surface horizons. A roughly static ratio throughout the pedon may indicate that weathering equals leaching.

As we hypothesized, the total Cae pool varies across the 11 pedons, but the distribution pattern of the soil Cae pool with-in the pedons also varies with depth. Parent material variation across serpentinitic landscapes was a factor in Ca pool variations

Table 4. Calcium and Mg content of parent material (Capm and Mgpm), fi ne earth (soils) total (Cast and Mgst) and extractable (Cae and Mge), and leaves of Vulpia microstachys (Nutt.) Munro plants (Cap and Mgp) grown in 500-cm3 containers fi lled with soil material from Henneke soil series map units.

Survey area Mgpm Capm Mgst Cast Mge Cae Mgp Cap Capm/Mgpm Cast/Mgst Cae/Mge Cap/Mgp

—––———————–g kg−1——————–—Colusa County 85 5.8 92 14.7 2.0 0.5 2.4 2.4 0.07 0.16 0.24 1.02

Glenn County 196 1.0 133 2.6 3.0 0.8 2.1 1.9 0.01 0.02 0.26 0.92

Kings County 150 2.3 41.8 14.1 2.8 2.6 4.2 3.5 0.02 0.34 0.94 0.84

Napa County 192 0.1 22.1 27.9 1.5 3.9 1.4 6.5 <0.01 1.26 2.56 4.67

San Benito County 212 10.3 211 4.9 0.6 0.2 0.6 0.7 0.05 0.02 0.26 1.19

Shasta County 200 0.7 114 15.2 1.4 0.6 2.2 2.0 <0.01 0.13 0.42 0.92

Tehama County 35 164 54.8 90 2.7 3.0 1.3 6.6 4.7 1.64 1.15 5.01

Mariposa County 221 0.4 104 13.8 5.0 1.2 1.4 3.1 <0.01 0.13 0.25 2.13

Mendocino County 155 9.4 163 9.6 2.6 0.7 6.1 1.5 0.06 0.06 0.26 2.49

Shasta-Trinity National Forest 188 0.1 100 6.0 3.0 0.7 4.5 2.3 <0.01 0.06 0.22 0.50Los Padres National Forest 222 0.1 167 2.3 1.5 0.4 1.2 1.6 <0.01 0.01 0.24 1.34

SSSAJ: Volume 73: Number 6 • November –December 2009 2093

laterally across the landscape and perhaps to some degree verti-cally within the pedon (Table 4).

Plant Response to Calcium and Magnesium PoolsTotal elemental contents of the container soils (all fi ne earth

fraction) for Napa and Tehama County yielded Cast/Mgst ratios >1 in agreement with the total elemental analysis by horizon for each pedon. Th e soil Cae/Mge ratio for the other container soils (Table 4) were also generally in agreement with the extractable cations from the pedons (Table 3).

We recognized that the parent material mineralogy was vari-able in soils formed on serpentinitic landscapes, and that parent material mineral content variability infl uenced the Cae pool. For data analysis, we grouped the soils into three groups based on parent material. Th e fi rst comprised the seven (n = 7) serpenti-nite-derived pedons that had no more than trace amounts of Ca-bearing minerals. Th e second comprised all 11 (n = 11) pedons, and the third (n = 9) consisted of the seven serpentinite-derived pedons that had no more than trace amounts of Ca-bearing min-erals plus the two pedons underlain by serpentinites that had mi-nor amounts of Ca-bearing minerals (Kings and Shasta counties).

Exchangeable Calcium and Magnesium Relation to Plant Calcium and Magnesium

For pooled data from all pedons (n = 11), the Cae/Mge ratio had a signifi cant positive correlation with leaf Cap/Mgp (r2 = 0.55, P < 0.01). For serpentinite-derived soils only (n = 9), the Cae/Mge ratio to Cap/Mgp ratio correlation was not signifi cant. Excluding serpentinite sites that contained only a minor amount of Ca-bearing minerals and the non-serpentinite sites improved the correlation, but it was not signifi cant (Fig. 2). Th erefore, including non-serpentinitic sites in the data set exerts enough infl uence to increase the strength of the correlation and add sig-nifi cance. Th e bimodal distribution, serpentinite vs. non-serpen-tinite, is clearly evident in Fig. 2 and underscores the variability of lithology on serpentinitic landscapes.

Fine Earth Calcium and Magnesium Relation to Plant Calcium and Magnesium

Similar to the Cae/Mge ratios, if the non-serpentinite sites are included (n = 11) the Cast/Mgst ratio analyses were numeri-cally highly correlated with the Cap/Mgp ratio (r2 = 0.81), but the correlation was not signifi cant. Also similar to the Cae/Mge ratio, the Cast/Mgst to Cap/Mgp ratio correlations were not sig-nifi cant for serpentinite-only sites, or for serpentinites without minor Ca-bearing minerals (Table 5).

Either the Cae/Mge ratio or the Cast/Mgst ratio helps sepa-rate non-serpentinite from serpentinite parent materials but fails in any predictive power to relate the Cap/Mgp ratio for serpenti-nites. Because low Ca availability is a primary limitation to veg-etative vigor, Cap content relationships are perhaps more useful than the Cap/Mgp ratio.

Fine Earth Calcium and Magnesium Relation to Plant Calcium

When including the non-serpentinite sites (n = 11), Cast was signifi cantly correlated to Cap (r2 = 0.64, P = 0.01); again we point out that a clearly bimodal distribution is evident when including the non-serpentinite sites. Removing the non-serpen-

tinite sites weakened the correlation, and it was no longer signifi -cant (Table 5). Removing non-serpentinites and serpentinites with minor amounts of Ca-bearing minerals resulted in a similar insignifi cant correlation (Table 5). Th erefore, we conclude that Cast is not a useful predictor of Cap for these soils.

High soil Mg contents may interfere with Ca uptake and therefore infl uence plant performance on serpentinite land-scapes. Indeed the Mgst was signifi cantly negatively correlated to Cap. Unlike the Cast relationships, however, removing the serpentinite sites without minor Ca-bearing minerals weakened the relationship to Cap but it remained signifi cant (Table 5). We discuss the Mgst negative correlation below.

Exchangeable Magnesium Relation to Plant Calcium While the Cae/Mge ratio, Cast/Mgst ratio, Cast, and Mgst

relationships with Cap are helpful to separate non-serpentinite- from serpentinite-derived soils, total elemental analysis (Cast and

Fig. 2. Correlations of soil extractable Ca/Mg ratio (Cae/Mge) to leaf Ca/Mg ratio (Cap/Mgp) in Vulpia microstachys (Nutt.) Munro: ♦ and X are non-serpentinite parent materials; 6 and 7 are serpentinite sites with minor amounts of Ca-bearing minerals; @ are serpentinite sites without minor amounts of Ca-bearing minerals. ** Signifi cant at the 0.01 probability level.

2094 SSSAJ: Volume 73: Number 6 • November–December 2009

Mgst) is not a routine analysis by soil scientists, and the soil ex-changeable pool (Cae and Mge) probably refl ects nutrients more readily available to plants. Unlike Mgst, Mge was not correlated with Cap (r2 = 0.02) even when including the non-serpentinite sites. When the non-serpentinite sites were removed, Mge had a signifi cant correlation to Cap (Table 5). When the serpentinites with minor amounts of Ca-bearing minerals were removed, Mge also had a signifi cant correlation to Cap (Table 5).

Exchangeable Calcium Relation to Plant CalciumAs expected, Cae was signifi cantly correlated to Cap

(r2 = 0.89, P < 0.001) when all 11 soils were considered. When non-serpentinites were removed, Cae remained signifi cantly cor-related to Cap. Removing serpentinites without minor amounts

of Ca-bearing minerals reduced the correlation, but it remained signifi cant (Table 5).

Implications of Serpentinite Variability for Plant Calcium Uptake

Th e Cae/Mge ratio, the Cae pool, the distribution of Cast and Cae with depth, and CEC all varied across soils formed on serpentinite landscapes. For these 11 pedons, the Cast/Mgst ra-tio clearly delineated the non-serpentinite-derived soils (Table 5), so there is value in using fi ne earth fraction total elemental analysis to separate non-serpentinite- from serpentinite-derived soils within the serpentinite landscapes. Within only serpenti-nite-derived soils, however, total elemental analysis Cast/Mgst or Cae/Mge ratios were of little value for discriminating among serpentinite-derived soils or for predicting plant performance on these soils.

Correlations of Vulpia microstachys Cap content were positive with Cae and Mge, but were negatively correlated with Mgst (Table 5). From our limited data set, the negative correlation with Mgst probably refl ects the low total Ca in the parent material and result-ing limited supply of plant-available Ca. Why was the Mge posi-tively correlated to Cap? As Mge increased, Cae also increased (Fig. 3). We noted that that Cae/Mge ratio (Table 2) was greater than the Cast/Mgst ratio (Table 3) for pedons with a low total Cast/Mgst ratio (Table 3). Th is was also evident in the blended solum container soil, as expected (Table 4). Th is leads us to speculate that biocycling plays some role in Ca retention in the soil. Furthermore, an increased pool size of extractable cations allowed the plants to exploit a larger Cae pool. Th e larger pool size eff ect was evident in the Kings County soils where greater CEC and greater fi ne earth fraction volume appears to have dampened the eff ects of Ca biocycling. In contrast to the poor predictive ability of using the Cast/Mgst and Cae/Mge ratios on serpentinite-derived soils, Cae proved to be a good predictor of plant performance. Parent ma-terial Ca content clearly contributed to the Cast content, and to the Cae. We stratifi ed the soils into three groups based on min-eralogy and Ca-bearing minerals: (i) serpentinite with no more than trace amounts of Ca-bearing minerals as detected by XRD, (ii) non-serpentinite, and (iii) serpentinite with minor amounts of Ca-bearing minerals as detected by XRD. Th ese three groups were related to the Cae/Mge ratio. Th e serpentinite with no more than trace amounts of Ca-bearing minerals, as detected by XRD, generally had a Cae/Mge ratio of 0.2 or less in some part of the profi le (n = 7). Th e non-serpentinite soils had a Cae/Mge ratio >1.0, as expected. Th e soils derived from serpentinite with minor amounts of Ca-bearing minerals, as detected by XRD, had Cae/Mge ratios from 0.2 to 1.0. Th ese groups allowed us to confi rm that even when the parent material is serpentinitic, the variability in Ca-bearing mineral amounts can aff ect the relative proportion of Cae. Th e Cae is presumed to be the pool that plants can utilize.

Our work shows that Cae is a better predictor of plant performance (as indicated by Cap content) than are Cast, the Cast/Mgst ratio, or the Cae/Mge ratio. Furthermore, we have demonstrated that Cast and Mgst values help separate serpenti-nite parent material from non-serpentinites. Finally, mineralogic variability in serpentinite parent materials probably translates into variability in plant-available Ca and probably accounts for much of the variation in plant viability or vigor.

Table 5. Selected correlations between soil and plants on ser-pentinitic landscapes.

Soil dependent variable† Plant variable‡ r2, n = 9§ r2, n = 7¶

Cast/Mgst, kg kg−1 Cap/Mgp, kg kg−1 0.03 0.04Cast, g kg−1 Cap, g kg−1 0.40 0.40

Mgst, g kg−1 Cap, g kg−1 −0.90*** −0.85**

Mge, g kg−1 Cap, g kg−1 0.45 0.72*

Cae, g kg−1 Cap, g kg−1 0.69** 0.66*

* Signifi cant at the 0.05 probability level.** Signifi cant at the 0.01 probability level.*** Signifi cant at the 0.001 probability level.† Cast, soil fi ne earth fraction total Ca; Mgst , soil fi ne earth fraction total Mg; Cae , extractable Ca; Mge, extractable Mg.‡ Cap, Ca content of Vulpia microstachys leaves; Mgp, Mg content of Vulpia microstachys leaves.§ Serpentinite-derived sites.¶ Serpentinite sites with no minor Ca-bearing minerals detectable by x-ray diffraction.

Fig. 3. Correlation of soil extractable Mg (Mge) to soil extractable Ca (Cae). *** Signifi cant at the 0.001 probability level.

SSSAJ: Volume 73: Number 6 • November –December 2009 2095

REFERENCESAlexander, E.B., R.G. Coleman, T. Keeler-Wolf, and S.P. Harrison. 2007.

Serpentine geoecology of western North America. Oxford Univ. Press, New York.

Alexander, E.B., W.E. Wildman, and W.C. Lynn. 1985. Ultramafi c (serpentinitic) mineral class. p. 135–146. In J.A. Kittrick (ed.) Mineral classifi cation of soils. SSSA Spec. Publ. 16. ASA and SSSA, Madison, WI.

Brooks, R.R. 1987. Serpentine and its vegetation: A multidisciplinary approach. Dioscorides Press, Portland, OR.

Carlson, R.M. 1978. Automated separation and conductimetric determination of ammonia and dissolved carbon dioxide. Anal. Chem. 50:1528–1531.

Coleman, R.G., and T.C. Keith. 1971. A chemical study of serpentinization: Burro Mountain, California. J. Petrol. 12:311–328.

Deer, W.A., R.A. Howie, and J. Zussman. 1997. An introduction to the rock-forming minerals. 2nd ed. Geol. Soc., London.

Gavlak, R.G., D.A. Horneck, and R.O. Miller. 1994. Plant, soil, and water reference methods for the Western Region. Western Rural Development Center, Corvallis, OR.

Gee, G.W., and J.W. Bauder. 1986. Particle size analysis. p. 383–411. In A. Klute (ed.) Methods of soil analysis. Part 1. Physical and mineralogical methods. Agron. Monogr. 9. ASA and SSSA, Madison, WI.

Gordon, A., and C.B. Lipman. 1926. Why are serpentine and other magnesian soils infertile? Soil Sci. 22:291–302.

Jenny, H. 1980. Th e soil resource. Springer-Verlag, Berlin.Kruckeberg, A.R. 1985. California serpentines: Flora, vegetation, geology, soils,

and management problems. Univ. of California Press, Berkeley.Loew, O., and D.W. May. 1901. Th e relation of lime and magnesia to plant growth:

I. Liming of soils from a physiological standpoint; II. Experimental study of the relation of lime and magnesia to plant growth. USDA Bur. of Plant Ind. Bull. 1. U.S. Gov. Print. Offi ce, Washington, DC.

McGahan, D.G. 2007. A survey of soils formed on serpentinitic landscapes in California. Ph.D. diss. Univ. of California, Davis.

McGahan, D.G., R.J. Southard, and V.P. Claassen. 2008. Tectonic inclusions on serpentinitic landscapes contribute plant nutrient calcium. Soil Sci. Soc. Am. J. 73:838–847.

Merrill, G.P. 1906. A treatise on rocks, rock-weathering and soils. Macmillan Co., New York.

O’Hanley, D.S. 1996. Serpentinites: Records of tectonic and petrological history. Oxford Univ. Press, New York.

Page, N.J. 1966. Mineralogy and chemistry of the serpentine group minerals and the serpentinization process. Ph.D. diss. Univ. of California, Berkeley.

Page, N.J. 1967. Serpentinization considered as a constant volume metasomatic process: A discussion. Am. Mineral. 52:545–549.

Pettijohn, F.J. 1941. Persistence of heavy minerals and geologic age. J. Geol. 79:610–625.

Proctor, J., and S.R.J. Woodell. 1971. Th e plant ecology of serpentine: I. Serpentine vegetation of England and Scotland. J. Ecol. 59:375–395.

Quideau, S.A., R.C. Graham, O.A. Chadwick, and H.A. Wood. 1999. Biogeochemical cycling of calcium and magnesium by ceanothus and chamise. Soil Sci. Soc. Am. J. 63:1880–1888.

Rabenhorst, M.C., and J.E. Foss. 1981. Soil and geologic mapping over mafi c and ultramafi c parent materials in Maryland. Soil Sci. Soc. Am. J. 45:1156–1160.

Sawhney, B.L., and D.E. Stilwell. 1994. Dissolution and elemental analysis of minerals, soils and environmental samples. p. 49–82. In J.E. Amonette and L.W. Zelazny (ed.) Quantitative methods in soil mineralogy. SSSA, Madison, WI.

Soil Survey Division Staff . 1993. Soil survey manual. Agric. Handb. 18. U.S. Gov. Print. Offi ce, Washington, DC.

Soil Survey Staff . 1999. Soil Taxonomy: A basic system of soil classifi cation for making and interpreting soil surveys. 2nd ed. NRCS, Washington, DC.

Soil Survey Staff . 2004. Soil survey laboratory methods manual. Soil Surv. Invest. Rep. 42, Ver. 4.0. Available at soils.usda.gov/technical/lmm/ (accessed 8 Feb. 2008; verifi ed 9 Aug. 2009). NRCS, Lincoln, NE.

Soil Survey Staff . 2007. Offi cial soil series descriptions. Available at soils.usda.gov/technical/classifi cation/osd/index.html (accessed 8 Feb. 2008; verifi ed 9 Aug. 2009). NRCS, Washington, DC.

Vlamis, J., and H. Jenny. 1948. Calcium defi ciency in serpentine soils as revealed by absorbent technique. Science. 107:549.

Walker, R.B., H.M. Walker, and P.R. Ashworth. 1955. Calcium–magnesium nutrition with special reference to serpentine soils. Plant Physiol. 30:214–221.