JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Burke, R. A.
Right arrow Articles by Piccolo, M. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Burke, R. A.
Right arrow Articles by Piccolo, M. C.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Burke, R. A.
Right arrow Articles by Piccolo, M. C.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Soil Microbiology
Right arrow Isotopes
Right arrow Soil Organic Matter
Journal of Environmental Quality 32:198-206 (2003)
© 2003 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Landscape and Watershed Processes

Stable Carbon Isotope Ratio and Composition of Microbial Fatty Acids in Tropical Soils

Roger A. Burke*,a, Marirosa Molinaa, Julia E. Coxb, Laurie J. Osherc and Marisa C. Piccolod

a USEPA, National Exposure Research Lab, 960 College Station Rd., Athens, GA 30605
b Dep. of Biological Sciences, Univ. of Arkansas, Fayetteville, AR 72701
c Dep. of Plant, Soil, and Environmental Sciences, Univ. of Maine, Orono, ME 04469
d Centro de Energia Nuclear na Agricultura (CENA-USP), Laboratorio de Biogeoquimica Ambiental, Av. Centenario 303 cep 13416.000, Piracicaba, SP Brasil

* Corresponding author (burke.roger{at}epa.gov)

Received for publication February 15, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The soil microbial community plays a critical part in tropical ecosystem functioning through its role in the soil organic matter (SOM) cycle. This study evaluates the relative effects of soil type and land use on (i) soil microbial community structure and (ii) the contribution of SOM derived from the original forest vegetation to the functioning of pasture and sugarcane (Saccharum spp.) ecosystems. We used principal components analysis (PCA) of soil phospholipid fatty acid (PLFA) profiles to evaluate microbial community structure and PLFA stable carbon isotope ratios ({delta}13C) as indicators of the {delta}13C of microbial substrates. Soil type mainly determined the relative proportions of Gram positive versus Gram negative bacteria whereas land use primarily determined the relative proportion of fungi, protozoa, and actinomycetes versus other types of microorganisms. Comparison of a simple model to our PLFA {delta}13C data from land use chronosequences indicates that forest-derived SOM is actively cycled for appreciably longer times in sugarcane ecosystems developed on Andisols (mean turnover time = 50 yr) than in sugarcane ecosystems developed on an Oxisol (mean turnonver time = 13 yr). Our analyses indicate that soil chronosequence PLFA {delta}13C measurements can be useful indicators of the contribution that SOM derived from the original vegetation makes to continued ecosystem function under the new land use.

Abbreviations: G-, Gram negative • G+, Gram positive • GC, gas chromatography • FAME, fatty acid methyl ester • MS, mass spectrometry • PC, principal component • PCA, principal components analysis • PLFA, phospholipid fatty acid • SOM, soil organic matter


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
TERRESTRIAL ECOSYSTEM functioning is largely governed by the soil microbial community through soil organic matter (SOM) decomposition and nutrient cycling (Kennedy and Smith, 1995). Soil microbial populations also play an important role in the global budgets of the key gases carbon dioxide, nitric oxide, and nitrous oxide (Paul and Clark, 1996). Therefore, microbial community responses to disturbances such as land use change could have major influences on carbon and nutrient dynamics and trace gas exchange. The actual role of soil microbes in these processes is difficult to evaluate, however, because little is known about the structure and diversity of soil microbial communities.

A necessary step in developing our knowledge of soil microbial ecosystems is to identify suitable indicators of the microbial communities and SOM cycling. Biomarker compounds, such as phospholipid fatty acids (PLFAs), are one type of microbial biomass and community structure indicator (Tunlid and White, 1992). Phospholipids are found in all bacterial cells in reasonably constant amounts under steady state conditions, are rapidly lost from viable cells upon death, and have a high natural turnover rate; therefore, their abundance provides a useful measure of viable or potentially viable microbial biomass (White, 1993). In some cases (e.g., methanotrophs), PLFA composition is indicative of a specific microbial group (Ringelberg et al., 1989). Generally PLFA data are useful in statistical tests of broad community structure differences among different ecosystem types or soil types or among ecosystems with different disturbance histories (Baath et al., 1995; Zelles et al., 1994, 1995a,b). Soil PLFA measurements have been made in relatively few ecosystems, however, and are especially scarce in tropical ecosystems.

Tropical ecosystems exchange more carbon with the atmosphere on an annual basis than ecosystems in any other region (Box, 1988). Further, tropical ecosystems are currently undergoing rapid changes in land use and land cover that often lead to net transfer of carbon to the atmosphere (Davidson and Ackerman, 1993). The nature of C dynamics in tropical regions is therefore of prime importance to global change research. Unfortunately, assessing and modeling the contribution of soils to the C cycle is complicated because SOM consists of many different fractions, which range from nearly inert (turnover times of thousands of years) to highly labile (e.g., microbial biomass), with widely varying decomposition rates.

In an effort to unravel this complex soil C cycle, researchers have used comparisons of the stable carbon isotopic composition ({delta}13C) of SOM before and after land use conversion. Natural abundance 13C/12C measurements are most effective in ecosystems in which an intentional change in land use at a known time caused a complete shift in photosynthetic pathway of the plant community, that is, from C3 ({delta}13C range from -33 to -22{per thousand}) to C4 ({delta}13C range from -16 to -9{per thousand}) or C4 to C3. Most studies have employed natural abundance 13C/12C measurements of bulk SOM to study the influence of land use change on carbon dynamics (Balesdent et al., 1987; Bernoux et al., 1998; Cerri et al., 1985; Townsend et al., 1995; Vitorello et al., 1989). The {delta}13C of bulk SOM represents a combination of inert and labile material, however, whereas the {delta}13C of PLFAs reflects only the material metabolized by soil microbes. The {delta}13C values of individual PLFAs indicate the source (e.g., C3 vs. C4 plant material) used by the active soil microbial community and thus, in conjunction with the known vegetation history, yield information about the SOM cycling rate.

The effects of land use, soil type, and climate regime on the size and structure of microbial communities and carbon dynamics in terrestrial ecosystems are major unknowns in global carbon cycle assessments. To address this uncertainty, we used measurements of the composition and 13C/12C of soil PLFAs, in combination with the concentration and 13C/12C of bulk SOM, in selected native forest, sugarcane, and pasture ecosystems from three distinct tropical regions. The 13C/12C of soil PLFAs should be a better indicator of the carbon source utilized by microbial communities than the {delta}13C of CO2 emitted from the soil surface (Feigl et al., 1995) because the latter measurement includes contributions from plant respiration. Measurement of the {delta}13C of in situ PLFAs extracted from soil should also more accurately reflect in situ microbial metabolism than measurement of the {delta}13C of CO2 produced in laboratory incubations of sieved soil (Townsend et al., 1995).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sampling Sites
Hawaii
Study sites were located in two areas of the Hamakua Coast of Hawaii on the northeast flank of Mauna Kea volcano at an elevation of about 700 m (see Osher [1997] for detailed description). Soils of the two areas, which are classified as Typic Distrandepts (Townsend et al., 1995), developed on similar volcanic ash deposits, estimated to be 15 000 to 20 000 yr old, but have experienced different amounts of rainfall. Mean annual rainfall rates are about 4000 and 2500 mm, respectively, at the Humuula and Kalopa sites. Mean annual temperature is about 22°C in both areas. The dominant vegetation at the Kalopa pasture site is a mixture of C4 grasses, 70% a Digitaria sp. and 30% broomsedge bluestem (Andropogon virginicus L.). The Kalopa pasture site has been maintained in that land use for about 90 yr. Sugarcane, which is a C4 plant, has been grown at the Humuula site for about 90 yr and at the Kalopa site for about 50 yr. Typical management of sugarcane in this area includes fertilization with 85, 75, and 110 kg ha yr-1 of N, P, and K, respectively, liming every two years, and tilling to 40 cm every 4 to 6 yr (Bashkin and Binkley, 1998). The forest sites in both areas consist of native ‘ohi’a lehua (Metrosideros polymorpha Gaud.)–dominated rain forest using the C3 photosynthetic pathway. The mean {delta}13C values of major plant carbon sources to the soil microbial community are: forest litter -29.3{per thousand}, sugarcane tissues -12.1{per thousand}, and pasture plant tissues -11.1{per thousand} (Osher, 1997). In July 1997 eight soil cores were collected from each site, divided into 0- to 10- and 10- to 20-cm intervals, and composited. At the Kalopa pasture site a separate set of eight soil cores was collected from a different part of the pasture a week later and treated in the same manner as the other samples. The Kalopa pasture results are reported as an average of the results obtained from the two sets of samples.

Brazil
The sampling sites were located in a region with extensive sugarcane cultivation near the Centro de Energia Nuclear na Agricultura (CENA) of the University of São Paulo in Piracicaba, São Paulo at an elevation of about 575 m. Mean annual rainfall is about 1400 mm and the annual mean temperature is about 20°C. The soil is classified as a Dark Red Latosol in the Brazilian soil classification system and as a clayey, kalonitic, isothermic Typic Haplorthox in the U.S. soil classification system (Vitorello et al., 1989). The soils contain between 1 and 3.5% organic carbon in the upper 10 cm (Vitorello et al., 1989). All sites were located on a flat area and the maximum distance between them is 250 m (Vitorello et al., 1989). Samples were collected in April 1998 and came from a primary forest and from fields converted to sugarcane (cultivar RB 785148) 27 and 65 yr in the past. Aboveground and belowground production are approximately equal in both cultivated areas and average 9 x 104 kg ha-1 yr-1 (Vitorello et al., 1989). Prior to a recent ban, fields were typically burned each year after harvest and about 5 x 103 kg ha-1 of residue was left. Chemical fertilization of both sugarcane areas consisted of 350 kg ha-1 of 0–13–8 (N–P–K) fertilizer in the furrows at planting, 300 kg ha-1 of 12–0–30 (N–P–K) during initial growth, and 350 kg ha-1 of 12–0–30 (N–P–K) to the ratoon (regrowth after harvest without reseeding) 3 yr after planting (Vitorello et al., 1989). Every 5 yr the fields are plowed to a depth of 20 to 30 cm and replanted. The mean {delta}13C values of major plant carbon sources to the soil microbial community are: forest litter -26.4{per thousand}, sugarcane leaves -13.2{per thousand}, and sugarcane roots -12.8{per thousand} (Vitorello et al., 1989). Samples from the intervals 0 to 10 and 10 to 20 cm were collected from the walls of a pit dug at each site.

Ecuador
The sampling sites were located at an elevation of about 1400 m on the western slope of the Ecuadorian Andes in tropical lower montane forest on slopes of 10 to 15 degrees. Mean annual rainfall is 3200 mm and average daily minimum and maximum air temperatures are 17 and 26°C, respectively (Rhoades et al., 2000). The soils developed from andesitic volcanic ash deposited 2500 yr ago and are classified as Andic Humitropepts (Rhoades et al., 2000). In November 1998 four soil samples were collected about 5 to 10 m apart from each of two forest sites, two sugarcane sites, and two African bristlegrass [Setaria sphacelata (Schumacher) Moss] pasture sites from the same area as studied by Rhoades et al. (2000). The forest and pasture sites were located within an approximately 1-km2 area of the Maquipucuna Forest Reserve and the sugarcane plots were located about 5 km from the reserve. The sugarcane and pasture sites were all converted to their present land uses about 25 yr ago. The sugarcane fields that we sampled were not intensively managed as they were not fertilized or burned and were hand-planted at a low density and hand-harvested (Rhoades et al., 2000). The forest patches were at least 25 yr old with a closed canopy and a healthy, solid understory of ferns and forbs. The mean {delta}13C values of major plant carbon sources to the soil microbial community are: forest litter -29.1{per thousand}, sugarcane litter -13.9{per thousand}, and bristlegrass litter -11.5{per thousand} (Rhoades et al., 2000). Four samples each from the 0- to 5- and 5- to 10-cm depth intervals were collected at each sampling location and were composited before analysis.

Phospholipid Fatty Acid Analysis
The soil samples were frozen within a few hours of collection, shipped frozen to Athens (GA), and maintained at -70°C until they were freeze-dried. After freeze-drying the soils were sieved (2 mm) to remove coarse fragments and roots. Phospholipid fatty acids were extracted from the freeze dried soils by the Bligh–Dyer technique (Bligh and Dyer, 1959), purified with silica and aminopropyl solid phase extraction (SPE) columns, identified, and quantified by gas chromatography (GC) and GC–mass spectrometry (MS) techniques as described by Zelles (1999). Briefly, 50 g of dry soil was added to about 240 mL of a single phase mixture of methanol–chloroform–phosphate buffer in a 2:1:0.8 ratio in a teflon bottle and rotated in a roller mill (US Stoneware, East Palestine, OH) for two hours. The solvent mixture was decanted, extra volumes of chloroform and phosphate buffer were added, and the mixture separated into two phases overnight in a separatory funnel. Lipids collected in the organic phase were loaded onto an activated silica gel column (BondElut; Varian, Palo Alto, CA) and fractionated into neutral lipids, glycolipids, and phospholipids with volumes of chloroform, acetone, and methanol (Zelles and Bai, 1993). Phospholipids were methylated by mild alkaline methanolysis (White et al., 1979) to form fatty acid methyl esters (FAMEs). This methylation procedure is suitable only for transesterfication and will neither methylate free fatty acids nor destroy the cyclopropane ring of cyclopropane fatty acids (Grogan and Cronan, 1997). Fatty acid methyl esters were further purified with NH2 aminopropyl columns (BondElut) (Zelles and Bai, 1993) and dimethyl disulfide derivatives were formed to determine double bond position in monounsaturated fatty acids (Nichols et al., 1986).

Lipid samples were analyzed with a Hewlett-Packard (HP; Palo Alto, CA) 6890 Series GC equipped with a flame ionization detector (FID) and a 50-m DB-5 capillary column (film thickness = 0.33 µm, internal diameter = 0.2 mm; J&W Scientific, Folsom, CA). Individual compounds were quantified by FID response relative to an internal standard (20:0 ethyl ester) added prior to GC analysis and identifications were assigned based on relative retention times compared with a prepared standard mixture (Sigma, St. Louis, MO; and Matreya, Pleasant Gap, PA) that was run with each batch of samples. Subsequent analysis by GC–MS was used to verify compound identification and to identify FAMEs not in the standard mixture. The same temperature program and similar 30-m DB-5 column (film thickness = 0.25 µm, internal diameter = 0.25 mm) as used in the GC–FID instrument were used in an HP 5890 Series II GC interfaced to an HP 5972 mass selective detector. The mean CV associated with quantifying individual PLFAs ranged from 2.4 to 5.7%. Total PLFA content was calculated as the sum of all identifiable PLFAs. The mean CV associated with quantifying total PLFA was approximately 8%.

Fatty Acid Notation
Fatty acid notation consists of the number of C atoms followed by a colon and the number of double bonds present in the molecule. The position of the first double bond relative to the aliphatic end of the molecule follows the symbol omega, {omega} (e.g., 16:1{omega}7). Prefixes "i" (iso) and "a" (anteiso) represent the location of a methyl branch one or two carbons, respectively, from the aliphatic end (e.g., i15:0). Methyl branching in other positions is noted by the appropriate number from the aliphatic end as a prefix (e.g., 10Me16:0) or by the prefix "br" if the position is unknown. Cyclopropane fatty acids are designated by the prefix "cy" (e.g., cy19:0).

Statistical Analysis
We chose 28 PLFAs present in nearly all of the samples analyzed and representing a broad spectrum of taxonomic groups for analysis of microbial community composition. Fatty acid methyl ester concentrations were converted to mole percentage PLFA after appropriate mass balance correction for the added methyl group. All data were tested for normality and equal variance before further statistical analysis. A few of the PLFAs required either log-normal or inverse transformation to meet normality and/or equal variance assumptions. Microbial community structure differences were evaluated by principal components analysis (PCA). One-way analysis of variance (ANOVA) was performed on the principal components and other parameters and either the Tukey–Kramer or Fisher's LSD multiple-comparison test was used to test for differences at the P < 0.05 level. All statistical analyses were run with the NCSS 2000 software package (Number Cruncher Statistical Systems, 2000).

Stable Carbon Isotope Analysis
Stable carbon isotope ratios ({delta}13C) were determined by continuous flow isotope ratio mass spectrometry (IRMS) techniques. The {delta}13C of individual FAMEs was measured by GC–IRMS in at least duplicate and the {delta}13C of bulk SOM was measured by elemental analyzer (EA)–IRMS in at least duplicate by previously described techniques (O'Malley et al., 1997). Briefly, {delta}13C of individual PLFAs was determined with an HP 5890 Series II GC coupled through a combustion interface and cryogenic water trap to a Fisons Optima IRMS (Micromass UK Ltd., Manchester, UK). The GC system is also coupled to a Fisons MD 800 quadrupole MS with National Institute of Standards and Technology (NIST) and Wiley mass spectral databases (John Wiley & Sons, New York, NY) for compound identification. The GC system was fitted with a 60-m DB-5MS capillary column (film thickness = 0.25 µm, internal diameter = 0.32 mm). The oven temperature program was the same as for FAME analysis. Results of 13C/12C measurements are reported as per mil ({per thousand}) difference relative to the Pee Dee Belemnite (PDB) standard in the {delta} notation:

The mild alkaline methanolysis converting phospholipids to FAMEs cleaves the fatty acids from the glycerol backbone and introduces a methyl group into each PLFA that was not initially present. This methyl group contributes to the carbon isotopic composition of the FAME and must be accounted for to determine the true {delta}13C of the PLFA. This was accomplished by simple mass balance as described by Abraham et al. (1998). Because there is little or no isotopic fractionation associated with this type of derivatization (Rieley, 1994), we assume that the {delta}13C of the added methyl group is the same as that of the methanol (-49{per thousand}) used in the methylation reaction. The mean precision (one standard deviation) of the bulk SOM {delta}13C analyses as measured by EA–IRMS was ±0.1{per thousand}. The mean precision of the {delta}13C of well-resolved individual compounds in standard mixtures as measured by GC–IRMS was ±0.4{per thousand}. Due mostly to co-elution of compounds and uncertainties involved with background correction, the mean precision of the {delta}13C of individual FAMEs from the samples was ±0.7{per thousand}.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil Microbial Biomass Content, Percent Soil Organic Carbon, and Stable Carbon Isotope Ratio of Soil Organic Matter
Mean total PLFA content, percent organic C, and {delta}13C-SOM for the soils are presented in Table 1. Microbial biomass levels in the upper 20 cm of the sugarcane soils from Brazil and Hawaii are lower compared with the original forest as indicated by total PLFA content, although the differences are not statistically significant. Regardless of land use, the Ecuador sites all have similar soil PLFA levels. The Hawaii soils have significantly more organic C (P < 0.0001) than the soils from the other regions and the Hawaii forest soils have significantly more organic C (P < 0.015) than the Hawaii sugarcane soils. None of the other differences in organic C level with respect to management or region are significant. Working in nearby sites in Ecuador, Rhoades et al. (2000) also found no significant difference in soil organic carbon between sugarcane and pasture soils, but found that forest soils contained significantly more carbon than either of the agricultural land uses. The organic C levels that we measured in the Ecuador forest sites are considerably lower than the levels (approximately 6.5%) found by Rhoades et al. (2000) for similar forest sites from the same area. These differences probably reflect natural spatial variations. The {delta}13C-SOM from all of the forest sites is strongly 13C-depleted (-27.2 to -25.5{per thousand}) and reflects exclusive input of C3 vegetation. The SOM in the sites that have undergone a vegetation change from C3 to C4 has intermediate {delta}13C values (-24.9 to -17.8{per thousand}) reflecting a mixture of the two vegetation types. As would be expected, the SOM from the sites that have been in C4 vegetation for longer time periods is less 13C-depleted. The SOM from the 90 yr old Hawaii pasture site is slightly less 13C-depleted (by up to 3.5{per thousand} in the shallow layer) than the SOM from the 90 yr old Hawaii sugarcane field, probably reflecting greater residue inputs and less physical disturbance under pasture management.


View this table:
[in this window]
[in a new window]
  		Table 1. Mean total phospholipid fatty acid (PLFA) content, total organic C, and {delta}13C of total soil organic matter (SOM) for the soils.

 
Microbial Community Composition
The first principal component (PC1) of the PLFA values distinguishes the Brazil Oxisol from the Hawaii and Ecuador Andisols (Fig. 1) . At most sites, PC1 also differentiates the samples collected from the two different depths with the shallower sample always having the smaller value of PC1 (Fig. 1). The PLFAs that contribute significant positive loadings to PC1 are primarily branched chain saturated compounds that are generally considered to indicate Gram positive (G+) bacteria (Zelles, 1999). Indicators of Gram negative (G-) bacteria, primarily monounsaturated and cyclopropyl saturated fatty acids (Zelles, 1999), are responsible for the largest negative loadings on PC1. Phospholipid fatty acid PC2 differentiated land use, which can be seen in Fig. 1 as a gradual increase in PC2 from forest (black symbols) to pasture (gray symbols) to sugarcane (open symbols). There is some PC2 overlap in the different land use treatments, particularly in the case of the Ecuador samples, but in all cases PC2 for the sugarcane treatment is larger than PC2 for the corresponding forest site (Fig. 1). In the Hawaii (Kalopa) and Ecuador areas the pasture PC2 value is intermediate between the forest and sugarcane treatments although the differences are more pronounced for the Hawaii samples. Phospholipid fatty acid markers for fungi (18:2{omega}6), protozoa (20:4{omega}6), and the actinomycetes group of the G+ bacteria (10Me18:0) contributed significant positive loadings to PC2, whereas straight chain saturated fatty acids, which are general bacterial markers, and a few miscellaneous indicators of G+ and G- bacteria (Zelles, 1999) contributed significant negative loadings to PC2. Principal Components 1 and 2 explained 35 and 20% of the total PLFA variability, respectively.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 1. Plot of the first and second principal components (PCs) extracted from principal components analysis (PCA) of the phospholipid fatty acid (PLFA) data from all soil samples. HF = Humuula (squares), Hawaii forest; HC90 = Humuula 90-yr-old sugarcane field; KF = Kalopa (diamonds), Hawaii forest; KP = Kalopa pasture; KC50 = Kalopa 50-yr-old sugarcane field; BF = Brazil (circles) forest; BC = Brazil sugarcane fields; EF = Ecuador (triangles) forest; EC = Ecuador sugarcane fields; EP = Ecuador pastures; G+ = Gram positive bacteria; G- = Gram negative bacteria; B = bacteria; F = fungi; P = protozoa; and A = actinomycetes. Symbol labels are: D = deeper sample; S = shallower sample; A and B in front of the depth designation correspond to the two replicates for the Ecuador ecosystems (triangles); and numbers in front of the depth designations indicate the ages of the Brazil sugarcane fields (open circles). Symbol fills: forest = black; pasture = gray; and sugarcane = open.

 
To further evaluate the effects of soil type or region and land use on microbial community composition we pooled the two Ecuador sampling intervals at each site into one 0- to 10-cm sample, compared them with the 0- to 10-cm samples from the other two regions by PCA, and then performed analysis of variance on the means of the resulting principal components. This analysis (Table 2) further supports the above results that soil type or region (PC1) has a significant effect on microbial community structure (G+ vs. G- bacteria), and less strongly, that land use (PC2) significantly influences microbial community composition (fungi, protozoa, and actinomycetes vs. other microbes).


View this table:
[in this window]
[in a new window]
  		Table 2. Region and management effects on principal component (PC) mean scores, 0- to 10-cm interval. Differences based on analysis of variance and Fisher's LSD multiple-comparison test.

 
Phospholipid Fatty Acid Stable Carbon Isotope Ratios
Mean PLFA stable carbon isotope ratios are plotted versus time since conversion in Fig. 2 and versus {delta}13C-SOM in Fig. 3 . The PLFAs from all of the forest soils are strongly 13C-depleted and reflect exclusive use of C3 carbon by the soil microbial community with little isotopic fractionation (Fig. 2 and 3). For sites that have been in C4 vegetation for long periods (>60 yr), {delta}13C-PLFA are much closer to the {delta}13C of the C4 vegetation than to {delta}13C-SOM as indicated by the fact that these samples plot well above the 1:1 line in Fig. 3. We interpret this to indicate that the microbial communities are primarily using C4 carbon with little isotopic fractionation. For the areas with paired sugarcane fields and pastures of the same conversion age (Hawaii and Ecuador), PLFAs from the pasture sites show slightly greater (0.5–2.4{per thousand}) incorporation of 13C than PLFAs from the sugarcane sites (Fig. 2). As with the SOM, the {delta}13C of PLFAs from sites that have been in C4 vegetation for intermediate lengths of time are intermediate between values characteristic of C3 and C4 vegetation.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 2. Plot of mean soil phospholipid fatty acid (PLFA) {delta}13C for the upper 10 cm versus time since ecosystem conversion. Error bars are ± one standard deviation. Abbreviations, symbols, and fills the same as in Fig. 1. The equation was fit to the forest and sugarcane data only. Values of a in the equation are -13{per thousand} and -14{per thousand} for the Oxisol and Andisols trends, respectively. The numbers on the figure are values of k in the equation used to generate the Oxisol and Andisols curves.

 


View larger version (19K):
[in this window]
[in a new window]
  		Fig. 3. Plot of mean soil phospholipid fatty acid (PLFA) {delta}13C versus mean soil organic matter (SOM) {delta}13C for samples from the upper 10 cm. Abbreviations, symbols, and fills the same as in Fig. 1. Error bars are ± one standard deviation. 1:1 line shown for reference.

 
To evaluate how quickly the soil microbial community in these systems shifts from C3 to C4 carbon after land use change we compared the output of one-parameter models to the sugarcane PLFA {delta}13C data. We base the assumption that the soil microbial community is utilizing one SOM pool on a previous analysis of one- and two-compartment models of SOM cycling (Bernoux et al., 1998), which concluded that in the absence of substantial justification for dividing SOM into different compartments based on lability, the simplest models possible should be used. This simple analysis suggests that, for the Brazil Oxisol sugarcane soil microbial community in the upper 10 cm, the mean turnover time (1/k) to switch from metabolizing only forest-derived carbon to only sugarcane-derived carbon is about 13 yr (Fig. 2). For the sugarcane fields on volcanic soils from Ecuador and Hawaii, a similar analysis suggests that the mean turnover time for the microbial community to switch from C3 carbon to C4 carbon is about 50 yr (Fig. 2).

The {delta}13C values of selected PLFAs that are commonly accepted prokaryotic biomarkers are shown in Fig. 4 . Monounsaturated fatty acids (e.g., 16:1{omega}7) are generally indicative of G- bacteria, and iso or anteiso branched chain fatty acids (e.g., i15:0) in soils typically indicate G+ bacteria (Zelles, 1999). In the samples analyzed in the present study, i15:0 was generally the least 13C-depleted PLFA (Fig. 4). Fatty acids with methyl branching on the 10th carbon (e.g., 10Me16:0) are indicative of the actinomycetes group of G+ bacteria (Zelles, 1999). The 10Me16:0 was generally slightly more 13C-depleted than the i15:0 (Fig. 3). Cyclopropyl fatty acids (e.g., cy19:0) are most commonly attributed to G- bacteria but are also found in a few G+ genera (Zelles, 1999). The relative concentration of cy19:0 has been observed to increase under conditions of low oxygen level, stress, or starvation by various investigators (Grogan and Cronan, 1997; Guckert et al., 1986; Petersen and Klug, 1994). In the samples we analyzed here, cy19:0 was always the most abundant PLFA (data not shown) and the most 13C-depleted PLFA (Fig. 4). The range in {delta}13C of PLFAs in individual samples was typically 5 to 6{per thousand} (Fig. 2).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. Comparison of {delta}13C of selected soil phospholipid fatty acids (PLFAs) from the upper 10 cm. Abbreviations are the same as in Fig. 1.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
A number of studies have demonstrated varying carbon isotopic fractionations by microbial species grown in laboratory settings. For instance, a group of bacterial and fungal species displayed a wide range (approximately 17{per thousand}) in isotopic fractionation between the PLFA 16:0 and a variety of simple substrates (Abraham et al., 1998). Little PLFA–substrate isotopic fractionation was observed when a bacterial species was grown on a variety of complex substrates but large fractionation occurred when the same organism was grown on the one-carbon compounds methane and methanol (Salata, 1999). A G- bacteria grown in culture on lactate exhibited much larger PFLA–substrate isotopic fractionation under anaerobic conditions (about -10{per thousand}) than aerobic conditions (about -1{per thousand}) (Teece et al., 1999). Similarly, Cifuentes and Salata (2001) reported that isotopic fractionation between the PLFA 16:0 and bulk organic matter averaged about 0{per thousand} for aerobic soils and about -10{per thousand} in anaerobic environments where methane production was likely. The small differences (within a few per mil) that we observed between the mean {delta}13C-PLFAs and the {delta}13C of the SOM or existing vegetation, combined with previous results (Cifuentes and Salata, 2001), suggest that when complex microbial communities utilize a complex mixture of substrates in aerobic soils, the different individual fractionations tend to cancel each other out. Thus, available evidence strongly suggests that mean PLFA {delta}13C values are good indicators of the mean {delta}13C of microbial substrate in predominantly aerobic soils.

Disturbance associated with conversion of native tropical forest to agricultural uses generally disrupts SOM cycling and frequently leads to a loss of fertility within a few years (Tiessen et al., 1994). Plant productivity and ecosystem function in many tropical soils that are poor in inorganic nutrients are dependent upon SOM cycling via microbial growth and death (Singh et al., 1989; Tiessen et al., 1994), and the length of time after conversion that forest-derived SOM is available to the microbial community is important to the continued fertility of the agricultural soil (Cerri et al., 1991). Comparison of a simple model to the PLFA {delta}13C of our chronosequence samples (Fig. 2) has shown PLFA {delta}13C data to be useful as indicators of the contribution of forest-derived carbon to the maintenance of agroecosystem function through time.

The 5 to 6{per thousand} range in PLFA {delta}13C values observed in most of these samples (Fig. 2–4) probably reflects the different isotope fractionations associated with the different fatty acid biosynthetic pathways responsible for the variety of fatty acids produced (Abraham et al., 1998; Teece et al., 1999). Although the greater fractionation against 13C observed under anaerobic conditions (Teece et al., 1999) does not greatly influence the mean {delta}13C-PLFA (Fig. 2–4) in the predominantly aerobic soils studied here, it could explain why the cy19:0 was always the most 13C-depleted PLFA if formation of cy19:0 is enhanced by anaerobic conditions (Guckert et al., 1986) that could occur within microsites in these soils. Thus, in addition to providing information about microbial substrates, {delta}13C values of individual PLFAs may be useful indicators of biosynthetic pathways in soils.

The slightly greater (0.5–2.4{per thousand}) {delta}13C in PLFAs from the Hawaii and Ecuador pasture sites as compared with the sugarcane sites of the same conversion age (Fig. 2) is undoubtedly due, in part, to the fact that the pasture grasses are slightly richer in 13C (Hawaii = 1{per thousand} [Osher, 1997], Ecuador = 2.4{per thousand} [Rhoades et al., 2000]) than the sugarcane. It was previously shown that sugarcane soils have less carbon than do pasture soils for these same Hawaii sites, however, probably due to a combination of lower carbon inputs and greater levels of soil disturbance resulting from cane cultivation (Osher, 1997). The stronger 13C depletion of the PLFAs and the lower total microbial biomass (Table 1) observed in the Hawaii sugarcane soil compared with the pasture of the same age (Fig. 2) suggest that the sugarcane soil microbial community may be adapted to more recalcitrant C3 substrates. Mineral–SOM interactions could also contribute to the pasture–sugarcane {delta}13C-PLFA difference. It has been argued that, compared with temperate soil, tropical soils tend to have a smaller labile SOM pool but a larger physically or chemically protected SOM pool that is highly vulnerable to release due to disturbance (Zech et al., 1997). Thus it is possible that disturbance associated with sugarcane cultivation has released previously protected C3 substrates, partially accounting for the slightly more 13C-depleted PLFAs in the Hawaii and Ecuador sugarcane soils compared with the pasture soils of the same conversion age.

Principal component analysis of our PLFA profiles indicates that microbial community structure varies measurably on different soil types and under different land uses. With respect to the soils in this study, overall soil type has a greater effect on microbial community structure than does management of that soil. A comparison of two California soils (a loam and a clay) subjected to various management parameters yielded similar conclusions based on PLFA profiles (Bossio et al., 1998). They attributed PLFA differences mainly to the clay soil being more anaerobic (Bossio et al., 1998). On the other hand, it was demonstrated that soil management had an influence on soil microbial community structure by comparing PLFA profiles in tropical forest and converted pineapple [Ananas comosus (L.) Merr.] plantation soils in Tahiti (Waldrop et al., 2000). Specifically, G+ biomarkers were relatively more abundant in forest soils than in plantation soils whereas markers for actinomycetes and fungi were relatively more abundant in the plantation soils (Waldrop et al., 2000). We observed a similar trend in that actinomycetes and fungi biomarkers are more important in our agricultural soils (Fig. 1). In contrast to the results of Waldrop et al. (2000), G+ markers tend to be relatively more abundant in our agricultural treatments compared with the corresponding forest sites (Fig. 1). In addition to the above PLFA studies, at least two investigations used genetic techniques to document differences in the soil microbial communities between adjacent tropical forest and pasture ecosystems (Borneman and Triplett, 1997; Nusslein and Tiedje, 1999). Nusslein and Tiedje (1999) calculated that conversion of forest to pasture in Hawaii caused a 49% change in soil microbial community composition, which they speculated was due to a change in the composition of the organic compounds input to the soil as a result of the change in plant community. We know of no other studies that have evaluated the effect of conversion of forest to sugarcane on soil microbial community structure.

Several studies have found that soil PLFA markers for G- bacteria increase with the availability of organic substrates (Bossio and Scow, 1998; Peacock et al., 2001), which is apparently related to the high intrinsic growth rates exhibited by many genera of G- bacteria (Atlas and Bartha, 1993). In contrast, many G+ bacteria are able to degrade complex substrates such as lignin and humic acid but have slower growth rates (Paul and Clark, 1996). Our PCA results (Fig. 1) indicate that G- markers tend to be higher (PC1 more negative) in forest soils compared with the corresponding agricultural treatments and G- markers are always higher in the shallower sample compared with the deeper one at a given site. These trends (Fig. 1) are generally consistent with the relationship between G- markers and substrate availability as carbon contents are generally higher in forest soils compared with adjacent agricultural soils and SOM and microbial biomass levels generally decrease with soil depth (Table 1; Nadelhoffer and Fry, 1988; Peacock et al., 2001). The trends of relatively more abundant G+ markers (PC1 more positive) in the agricultural soils and deeper samples are also consistent with the generally slower growth rates of G+ bacteria.

Our observation that PLFA and C levels are higher in the upper 10 cm of the Andisols from Hawaii as compared with the highly weathered Oxisol from Brazil, agrees with previous observations (Zech et al., 1997). The higher C levels in Andisols have been attributed to reduced rates of SOM decomposition due to SOM stabilization by physico–chemical interactions between SOM and the allophanic material (Zunino et al., 1982). Enhanced physico–chemical protection of labile SOM in the Andisols compared with the Oxisol may also explain why it takes so much longer (a factor of four times as long) for C4 carbon to replace C3 carbon as the primary microbial substrate in the Andisols (Fig. 2), as indicated by the mean turnover times calculated from sugarcane PLFA {delta}13C trends.

The higher C and PLFA levels observed in the Hawaii Andisols compared with those from Ecuador may be related to the respective soil ages, an estimated 15 000 to 20 000 yr old for the Hawaii soils (Osher, 1997) versus 2500 yr old for the Ecuador soils (Rhoades et al., 2000). There is evidence that it may take >6000 yr for young soils with little or no SOM to establish a stable carbon distribution (Schlesinger, 1990); thus the younger Ecuador soils may not have had sufficient time to reach their stable SOM and PLFA levels.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Phospholipid fatty acid {delta}13C values are reasonable indicators of the {delta}13C of microbial substrates in predominantly aerobic tropical soils. Comparison of simple models to PLFA {delta}13C values from a chronosequence of converted sites provides valuable indication of the contribution of SOM derived from the original vegetation to the maintenance of ecosystem function. Factors such as soil type or region and land use can have discernible effects on tropical soil microbial community composition. Principal components analysis of soil PLFA profiles suggests that G+ bacteria are relatively more important members of the microbial community of the Brazil Oxisol sites whereas G- bacteria are relatively more important in the Hawaii and Ecuador Andisol sites. Further, fungi, protozoa, and actinomycetes appear to be relatively more important members of the microbial community in the agricultural sites than in the forest habitats.


   ACKNOWLEDGMENTS
 
This research was supported by USEPA National Exposure Research Laboratory Internal Grant #AT15U1. We thank Dr. Greg Eckert of the U.S. National Park Service for providing the samples from Ecuador and Dr. Carlos Cerri of CENA-USP for helping us obtain the samples from Brazil. This paper has been reviewed in accordance with the USEPA's peer and administrative review policies and approved for publication. Mention of trade names or commercial products does not constitute an endorsement or recommendation for use by the USEPA.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Burke, R. A.
Right arrow Articles by Piccolo, M. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Burke, R. A.
Right arrow Articles by Piccolo, M. C.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Burke, R. A.
Right arrow Articles by Piccolo, M. C.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Soil Microbiology
Right arrow Isotopes
Right arrow Soil Organic Matter

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
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome