| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a College of Forest Resources, University of Washington, Box 352100, Seattle, WA 98195-2100 USA
b U.S. Geological Survey, Forest and Rangeland Ecosystem Science Center, Cascadia Field Station, Box 352100, Seattle, WA 98195-2100 USA
c School of Natural Resources, Soil Science, 302 ABNR Building, University of Missouri, Columbia, MO 65201 USA
sprich{at}u.washington.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
, while SW forest soils store more C than SW meadows
. Ecosystem C storage is greater in forests than in meadows. Under a warmer climatic scenario with drier summers and wetter winters, subalpine C storage may decrease in the NE and increase in the SW, and changes in C storage will be closely related to vegetation distribution, ecosystem productivity, decomposition rates, and local disturbance regimes. Because ecosystem processes and associated C storage differ between high- and low-elevation ecosystems, it is important that data from both high- and low-elevation sites are included in estimates of C storage in terrestrial ecosystems.
Abbreviations: CEC, cation-exchange capacity • dbh, diameter at breast height • LWD, large woody debris • NE, northeastern • SW, southwestern
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Soil C is an important component of C storage in most forest ecosystems (Vogt, 1991; Eswaran et al., 1993; Dixon et al., 1994). More than 50% of terrestrial C storage is contained in soils, and quantities are probably underestimated because many studies do not measure C storage in the lower soil profile (Eswaran et al., 1993; Kern, 1994). Accurate estimates of soil C distribution are further complicated by several factors, including high spatial variability in C content of soils, poor spatial coverage of areas with reliable estimates of C content in soils, and temporal variation in vegetation (Eswaran et al., 1993; Homann et al., 1998).
Subalpine ecosystems provide an instructive setting in which to study C dynamics. Mosaics of forests and meadows are characteristic of subalpine areas (Fig. 1) . Under global-warming scenarios (e.g., Gates et al., 1992), forest–meadow ecotones in subalpine regions are likely to shift in distribution. For example, it has been suggested that subalpine forests will expand as the climate becomes warmer (Zolbrod and Peterson, 1999). Altered distribution and quantities of high elevation C stores may accompany changes in forest–meadow ecotones. However, there are few data on C storage in these ecosystems, particularly regarding soils, on which to base predictions of responses to changes in local climate.
|
|
Materials and methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
|
|
|
|
Three forest soil profiles and three meadow soil profiles were described and sampled within each site; a total of 36 soil profiles were included in the study. Each soil pit was positioned where the ground surface was not noticeably disturbed and was excavated to bedrock or where rock fragments obstructed further digging. Sampling for physical and chemical analysis was conducted by genetic horizon.
In addition to mineral soil samples, organic horizon samples were obtained for each forest site. Sampling was conducted at six random locations within forest sites. A known volume was extracted for each sample with a 20-cm-diam. cylinder. Organic horizons in SW soils were sampled along with mineral soils. Organic horizons in NE meadow soils were absent.
Soil Analysis
All soil samples were analyzed for C, N, pH, exchangeable cations (Ca, Mg, K, Na), available P, cation-exchange capacity (CEC), and base saturation. Soil samples were air-dried, weighed, and sieved through a 2-mm screen prior to analysis. Organic horizon samples were air-dried and weighed prior to analysis. Subsamples of all mineral and organic soil samples were oven-dried at 105°C for 12 h to determine moisture corrections.
Subsamples of air-dried, <2-mm mineral soil were finely ground with a mortar and pestle and analyzed for C and N using a CHN analyzer (CHNS Analyzer Model 2400, Perkin-Elmer, Norwalk, CT). Organic horizon subsamples were analyzed in a similar manner but were ground with liquid N to ensure that they were fully homogenized.
Total soil C was calculated as a function of C concentration, bulk density, depth to bedrock, and rock volume (e.g., Homann et al., 1995). Large quantities of rock fragments made it difficult to accurately measure bulk density. Therefore, regression equations based on Homann et al. (1995) were used to calculate fine soil bulk density from C concentrations. Homann et al. (1995) developed regression equations for similar soil types in western Oregon with similar ranges in C concentration. Bulk density values of A and E horizons were replaced with measured bulk densities (Soil Survey Staff, 1992), which were more reliable in horizons near the surface than in deeper horizons. A fine-soil bulk density value of 1.56 g cm-3 was applied to all C horizons, after Homann et al. (1995). Soil C was assumed to be of organic origin in this study because soils are strongly acidic and carbonates are rarely present in parent materials (Tabor, 1975). We observed no free carbonates in any soil samples.
Soil pH was determined from 1:1 pastes of air-dried, <2-mm mineral soil in deionized water. Pastes were stirred and allowed to stand for 10 min to equilibrate prior to determination with a pH meter (PHM92 Lab pH meter, Radiometer, Inc., Westlake, OH).
Extractions for exchangeable-cation and CEC analyses were prepared concurrently. First, 5-g samples of air-dried, <2-mm mineral soil were saturated and extracted with 50 mL of 1.0 M NH4Cl for 12 h on a mechanical extractor. Solutions were collected and stored in a cold room prior to determination of exchangeable-cation concentration by inductively coupled Ar plasma emission spectroscopy (ICAP 61E, Thermo Jarrel Ash, Franklin, MA). The original soil sample was rinsed with ethanol to remove soluble NH+4. Soils were extracted with 50 mL of 1.0 M KCl for 12 h on a mechanical extractor to remove NH+4 from saturated exchange sites. Solutions were collected, stored in a cold room, and later analyzed for NH+4 concentration using a Technicon Autoanalyzer (Technicon Industrial Method no. 158-71W, 1977; Technicon, Tarrytown, NY) to determine CEC.
Available P was extracted using the Bray 1 available P method (Olsen and Sommers, 1982; Brown and Rodriguez, 1983). A soluble extraction was prepared by mechanically mixing 1 g of air-dried, <2-mm mineral soil with 10 mL of extracting solution for 10 min. The solution was then filtered and mixed with ascorbic acid molybdate solution. Available P was measured with a colorimeter.
Samples from prospective Bs and Bhs horizons (SW sites only) were analyzed for Fe and Al (Soil Survey Staff, 1992) to verify that samples met spodic horizon criteria. Two grams of air-dried, <2-mm soil were saturated with 200 mL of 0.2 M ammonium oxalate, and soil solutions were mixed on a mechanical shaker for 4 h in the dark. Solutions were then filtered through a Whatman no. 42 filter, and filtrates were analyzed for Al and Fe concentrations using inductively coupled argon plasma emission spectroscopy.
Biomass Sampling and Carbon Calculations
Above- and belowground biomass of forest components was calculated using allometric equations based on field measurements, and C content was estimated as 50% of biomass values. Field measurements were recorded within a slope-corrected 0.05-ha plot in each forest. Diameter at breast height (dbh), height, crown width, and crown height were measured for trees taller than breast height (1.37 m). Height and diameter of snags, as well as length, end, and midpoint diameter of logs >7.5 cm in diameter were recorded for measurements of large woody debris (LWD) within each plot. Aboveground and belowground biomass (i.e., coarse roots >5-mm diameter) were calculated with the program BIOPAK (Means et al., 1994), which uses allometric equations developed for Pacific Northwest tree species based on dbh and height. Understory forest vegetation below breast height was not included because it was very sparse in stands and comprised a very small proportion of the total biomass. The programs SNAG and LOG were used to calculate biomass of LWD (J. Means, 1994, personal communication). Fine woody debris biomass (debris <7.5-cm-diameter) was quantified using a line-intercept method (Brown, 1974). Bulk density values for fine woody debris were approximated as 0.48 g cm-3 for debris <2.5 cm in diameter and 0.40 g cm-3 for debris 2.5 to 7.5 cm in diameter (Harmon and Sexton, 1996).
Aboveground and belowground biomass within meadow plots was quantified by harvesting 0.25 by 0.25 m samples of aboveground meadow vegetation and sampling roots in the same dimension to the base of the maximum rooting depth, which ranged from 28 to 64 cm. Three samples were collected from each meadow. Aboveground samples included live and dead vegetation. Belowground samples included live roots, dead roots, and other organic matter. Roots were soaked in water to remove rock fragments and then washed over a 1-mm screen to remove residual soil. Aboveground vegetation samples and cleaned roots were then air-dried, weighed, ground in a Wiley mill and finely ground with a mortar and pestle using liquid N. Homogenized samples were analyzed for C content in a total CHN analyzer (Perkin-Elmer CHNS Analyzer Model 2400). Subsamples of aboveground vegetation and root samples were oven-dried at 105°C for 12 h to calculate moisture corrections.
Statistical Analysis
F tests were used to test for normality of data prior to subsequent analyses. Total soil C and ecosystem C were compared between all paired forests and meadows using paired t tests
. One-tailed t tests were included in the analysis because they determined whether C storage was greater in meadows vs. forests. Statistical results were considered nonsignificant at P > 0.10, marginally significant at 0.10 > P > 0.05, and significant at P < 0.05.
|
Results |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
|
Soil profile characteristics clearly differ among geographic locations and vegetation types (NE forests, NE meadows, SW forests, and SW meadows) (Table 2). Soils in both subalpine forests and meadows of the NE are classified as Lithic or Typic Dystrocryepts (Soil Survey Staff, 1998). Soil profiles have weakly expressed B horizons, but A horizons are rich in organic matter and contrast strongly with deeper mineral horizons. Surface horizons (A and BA) in NE meadows are thicker than NE forest A horizons. In some cases, surface horizons in NE meadows appear to have been thickened by mechanical mixing with B horizons during soil creep.
Soils in the SW subalpine areas are Lithic Haplocryods, Typic Humicryods, Lithic Humicryods, and Typic Humicryods. A few soil profiles in SW sites do not meet spodic horizon specifications and are classified as Lithic or Typic Dystrocryepts (Soil Survey Staff, 1998). Most SW soil profiles contain spodic horizons and thin E horizons, few Bhs horizons, and strongly illuvial Bs horizons with high concentrations of Fe and Al. Southwestern meadow soil profiles generally have less distinct horizons and weaker podzolization than SW forest soil profiles. Clay content is higher, and B horizons are more strongly expressed in SW soil profiles than NE soil profiles.
Soil Carbon and Chemical Properties
Soil Carbon Concentrations
Carbon concentrations of fine (<2 mm) mineral soil are high in upper horizons, ranging from 43 to 142 g kg-1 in NE soils and 27 to 162 g kg-1 in SW soils. Concentrations decrease with depth but are highly variable among soil profiles (Fig. 3)
. Soil C concentrations generally tend to be greater in NE meadows than NE forests, while concentrations in SW forests are often greater than those in SW meadows. These patterns are not evident within all soil profiles. Because horizon designation and depth vary considerably between sites, and particularly between NE and SW soils, statistical analyses were not conducted on C concentrations.
Total Soil Carbon
Total soil C (kg m-2) differs significantly between meadow and forest ecosystems. Northeastern meadow soils store more C than NE forests
, while SW forest soils store more C than SW meadows
. Although these comparisons generally are significant, soil C storage is highly variable within and among sites (Fig. 4)
.
|
|
, with large quantities of C in aboveground forest vegetation (Fig. 6)
. Northeastern forests store a majority of C in aboveground vegetation (
60%), with less than one-third of C contained belowground in coarse roots and mineral soil. Southwestern forests also store a majority of C in aboveground vegetation (65–70%), with approximately one-third of C stored belowground. Both forest types contain only 10% of ecosystem C in forest floor components (i.e., organic horizons, fine woody debris and large woody debris). Northeastern meadows store <5% of ecosystem C in aboveground vegetation, with 85% in mineral soil and 10 to 15% in roots. Southwestern meadows store 5% of C in aboveground vegetation, 55 to 60% of C in mineral soil, 15 to 25% in organic horizons, and 15% in roots.
|
|
Discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Soil Carbon and Chemical Properties
High soil C concentrations found in this study probably result in part from low soil temperatures, which can reduce rates of C mineralization. Carbon concentrations in fine soil fractions exceed 50 g kg-1 in several A and E horizons, and while C concentrations decrease sharply with depth, many BC and C horizons contain 5 to 10 g C kg-1 (Fig. 3). Previous studies of montane and subalpine soils have also documented high C concentrations (Sims and Nielson, 1986; Vogt et al., 1989).
Although C concentrations are high in the fine soil fraction, absolute quantities of C in the mineral soil profile are generally lower than in other Pacific Northwest forest ecosystems. Soil C storage in this study ranges from 3.5 to 14.5 kg m-2 and averages 8.6 kg m-2. In contrast, soil profiles sampled within a Pacific silver fir ecosystem in the Cascade Mountains of Washington (elevation 1140 m, 180-yr-old stand) contained 13.7 kg m-2 (Grier et al., 1981). A study of C storage in several montane forest soils in Oregon (at elevations considerably lower than our study) reports average values of 15.0 kg m-2 (Homann et al., 1995). Selected sites in low elevation, old-growth Douglas-fir [Pseudotsuga menziesii (Mirbel) Franco.] forests of the Oregon Cascades contained less soil C than in this study, with 4.5 to 6.5 kg m-2 (Grier and Logan, 1977). Finally, a study of low-elevation forests in Oregon and Washington reports mineral soil C values of 10.2 to 17.7 kg m-2 for Douglas-fir forests and 24.1 to 30.9 kg m-2 for western hemlock [Tsuga heterophylla (Raf.) Sarg.] forests <70 yr old (Edmonds and Chappell, 1994).
Several factors could have influenced absolute quantities of soil C in this study. Low total C storage results in part from the high volume rock fragments in soils (up to 75% in upper horizons, and 90% in deeper horizons), and the shallow, unstable slopes on which the study sites were located. Soils in the Pacific silver fir ecosystem studied by Grier et al. (1981) contain substantially fewer rock fragments and are located on gentle slopes, which may help explain the difference in C storage between the two studies. Homann et al. (1995) reported deeper soil profiles and lower rock fragments in their study of montane soils in Oregon. In addition, profile depths averaged 1.2 m in the study by Edmonds and Chappell (1994), compared with an average of 0.6 m in our study.
High soil C concentrations in fine mineral soil indicate a potential for much greater C storage than was reported in this study. High C concentrations in deeper horizons, such as Bhs and Bs horizons in SW soils, combined with the sheer volume of mineral soil in some soil profiles, can result in substantial quantities of deep C storage (Stone et al., 1993; Hammer et al., 1995; Richter et al., 1995). Some deposits of buried soils in cirque basins in the Olympic Mountains contain large quantities of C deep within the soil profile (Peterson and Hammer, 1994). Although soils on steep, planar slopes in subalpine areas of the Olympic Mountains may never reach potential C storage levels because they are shallow and skeletal, local convexities and toeslopes can amass large reserves of C that represent long-term storage (Richter et al., 1995; Christensen et al., 1999).
Because state factors, including parent material, topography and to some extent time of soil development, are similar among forest and meadow pairs, patterns of total soil C storage among study sites are probably related to ecosystem-level factors such as productivity, decomposition rates, and past disturbances. For example, high belowground primary productivity appears to have contributed to the high soil C storage in NE meadows. Northeastern meadows are dominated by graminoid species, which typically allocate large amounts of C to belowground biomass (Kuramoto and Bliss, 1970; Anderson and Coleman, 1985). Fine roots are abundant not only in upper horizons, but extend throughout deeper B horizons (Table 2a).
In contrast, SW forests store greater soil C than SW meadows, probably as a result of differences in overall productivity as well as accumulation of organic matter. Southwestern forests are more productive and accumulate greater organic matter in vegetation, detritus, and soils than SW meadows, which are not composed of graminoid species. In addition, the deep snowpack lasts longer in dense, cool forest stands of the SW and may promote soil organic matter accumulation by hindering decomposition (Sneddon et al., 1972).
Variability within soil profiles on shared planar slopes suggests that microtopographical features, such as bedrock undulations, local vegetation, and past soil disturbance unnoticeable at the surface, have affected soil development. High variability in soil profiles, including horizon thickness, profile depth, and rock fragment content, contribute, in turn, to within-site variability in total soil C.
Ecosystem Carbon
Ecosystem C estimates in this study are high compared with the few studies that have documented C storage in the Pacific Northwest (Table 4)
. Sites within this study were sampled above continuous treeline in a landscape mosaic of dense forests and small meadows. Therefore, C in aboveground and belowground storage varies spatially and does not extend throughout a continuous area.
|
Belowground biomass in forest soils probably was underestimated because fine roots were not sampled (Vogt et al., 1980).
Potential Changes in Soil Carbon in a Warmer Climate
The effects of climatic change may become evident within subalpine areas of the Olympic Mountains during the next century. Subalpine ecosystems in much of the Olympic Mountains are undergoing an observable shift in vegetation dominance (Woodward et al., 1995). Subalpine tree species, particularly in the SW Olympics and on north aspects in general, have established in subalpine meadows during periods of low snowpack during the last century (Kuramoto and Bliss, 1970; Woodward et al., 1995). If mean temperatures become warmer, as general circulation models predict for northern temperate latitudes (Gates et al., 1992), tree establishment will continue in many subalpine ecosystems (Rochefort et al., 1994). Changes in C storage in subalpine ecosystems will depend on whether climate becomes warmer and drier, or alternatively, warmer and wetter, and also on the particular region in which changes occur.
Forests in both the NE and SW store far more C than meadows. The greatest change in C storage in these ecosystems thus could be instigated by either expansion of forests into meadow areas (constituting a net sink of C) or conversion of forests to meadows through fire or gradual stand mortality (constituting a net source or loss of C). While differences in C storage between forests and meadows were found in soils, the major difference on an ecosystem level is in aboveground biomass. Therefore, while some soil properties may affect how much soil C is retained, accumulated, or lost in response to local warming, the main impact will be on the response of aboveground vegetation.
If climate becomes warmer and drier, subalpine ecosystems in the dry NE conceivably could lose C due to combined effects of decreased tree establishment and productivity in addition to more frequent fires. Tree establishment, already limited by summer drought, could cease altogether in NE meadows (Woodward et al., 1995). More frequent fires associated with summer drought could potentially expand meadow areas (Agee and Smith, 1984) and result in a substantial decrease in C storage. Alternatively, if climate becomes warmer and wetter, tree establishment in NE meadows could potentially constitute a long-term C sink.
A shift in local climate to warmer (and either drier or wetter) conditions in the SW could result in increased C storage. Tree establishment in meadows could continue and even increase if warmer conditions cause early melting or less accumulation of annual snowpack (Rochefort et al., 1994; Woodward et al., 1995).
Implications for Future Carbon Studies in Mountainous Systems
High variability in soils and among sites is an important feature of subalpine ecosystems. Resource managers and scientists who work in these systems should be aware that C pools in soils and vegetation are discontinuous, extremely variable, and subject to change through disturbance and local fluctuations in climate. In this study, soil C storage was often influenced by slight changes in microtopography, as evidenced by the high variability in C content among the three profiles of a given site. Other studies of mountainous systems have reported similar variability in soils (Grier, 1973; Holtmeier and Broll, 1992; Homann et al., 1995), and some soil chemical properties can vary seasonally in response to variation in temperature and precipitation (Weaver and Forcella, 1979). If C storage estimates are extrapolated to scales larger than those sampled in this study, it should be done cautiously and be accompanied by some quantification of variation in the soil and forest systems.
Because ecosystem processes and associated C storage can differ between high-elevation and low-elevation sites, it is important to include high-elevation C storage as a discrete component in C storage estimates for terrestrial ecosystems. Few studies in western North America have examined ecosystem C storage in subalpine areas (Grier et al., 1981; Boone et al., 1988; Gower and Grier, 1988; Sanscrainte, 1999). This study documents C storage in subalpine ecosystems of the Olympic Mountains, but it cannot represent the many conditions that exist in mountainous areas. Additional studies of C storage in high-elevation forests are necessary before C storage from mountainous regions in temperate biomes can be incorporated into temperate ecosystem C budgets with any confidence. Further study of the influence of site and soil properties on C storage and organic chemistry in subalpine ecosystems will also improve our understanding of how C storage at high elevations may be affected by potentially rapid climatic changes in the future.Sims Nielsen 1986
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 29, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. E. Bedison and A. H. Johnson Controls on the Spatial Patterns of Carbon and Nitrogen in Adirondack Forest Soils along a Gradient of Nitrogen Deposition Soil Sci. Soc. Am. J., October 21, 2009; 73(6): 2105 - 2117. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Osono and H. Takeda Fungal decomposition of Abies needle and Betula leaf litter. Mycologia, March 1, 2006; 98(2): 172 - 179. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Whitney and D. Zabowski Total Soil Nitrogen in the Coarse Fraction and at Depth Soil Sci. Soc. Am. J., March 1, 2004; 68(2): 612 - 619. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. B. Harrison, A. B. Adams, C. Licata, B. Flaming, G. L. Wagoner, P. Carpenter, and E. D. Vance Quantifying Deep-Soil and Coarse-Soil Fractions: Avoiding Sampling Bias Soil Sci. Soc. Am. J., September 1, 2003; 67(5): 1602 - 1606. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
