| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Ecosystem Science Division, ESPM, Univ. of California, Berkeley, CA 94720
b Soil Science Graduate Group, Land, Air and Water Resources, One Shields Ave., Univ. of California, Davis, CA 95616
c Geosciences Dep., The Pennsylvania State Univ., University Park, PA 16802
d Geography Dep., Univ. of California, Santa Barbara, CA 93106
e Soil Science Graduate Group, Land, Air and Water Resources, One Shields Ave., Univ. of California, Davis, CA 95616
* Corresponding author (rjsouthard{at}ucdavis.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
15N, DNA, C and N contents under mature shrubs and as a function of horizontal distance (0–3 m) away from the base of the shrubs. Shrub-base soils (0 m) contained more C and N, were slightly more acidic, and supported significantly larger microbial populations than soils between shrubs. The PLFA fingerprints also suggested that microbial communities, particularly at the shrub base, had a different composition than soils between shrubs, including a higher proportion of actinomycetes containing the biomarker 10me17:0. Soil respiration was generally highest at 0 m, corresponding with larger microbial biomass and larger C and N pools, but was highly variable, probably due to contributions from grasses and forbs. Average 15N values resembled plant material at the shrub base (4
) and were significantly isotopically enriched away from the shrubs (7), suggesting that fractionating losses of soil N occurred between shrubs. The elevated nutrient status of resource islands supported soil microbial communities that were larger, were different in character, respired more actively, and cycled N more tightly than those found in open spaces between shrubs. These open spaces "leak" isotopically light N from the soil.
Abbreviations: DOC, dissolved organic carbon • PLFA, polar lipid fatty acid
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Stable soil N isotopic signatures (15N) are a convenient tracer of net soil N-cycling effects because they represent the balance of biological N transformations (Amundson and Baisden, 2000). Previous work using soil 15N has suggested less conservative cycling of N in hot, dry locations, where losses due to isotope-fractionating processes result in higher soil 15N signatures than at cooler, more moist sites (Amundson and Baisden, 2000; Austin and Vitousek, 1998; Riley and Vitousek, 1995; Schulze et al., 1991). In arid zones, processes that favor loss of 14N (and thus increase soil 15N) may occur during brief episodes of denitrification following rainfall events; by NH3 volatilization, which is favored in dry, alkaline soils with low cation exchange capacity; or as loss of NOx and N2O during nitrification and denitrification (Mummey et al., 1997; Schlesinger et al., 1990; Smart et al., 1999). To the extent that N-cycling processes are a function of plant cover and climate, we hypothesized that these processes vary at the scale of microclimates beneath shrubs in desert ecosystems.
Elevated soil microbial biomass is commonly observed under shrub canopies (Bolton et al., 1993; Smith et al., 1994; Sarig et al., 1999), but variations in the character of the microbial community with shrub proximity have not been sufficiently explored. Microbial community profiles are likely to vary with shifts in microclimate and N-cycling processes between resource islands and open spaces (Steinberger et al., 1999; Kinsbursky et al., 1990; Herman, 1997). The aim of this study was (i) to determine whether variation in the soil microbial community coincides with variations in soil respiration and net N cycling between resource islands and shrub interspaces, and (ii) to provide a preliminary characterization of the nature of any observed variation in the microbial community.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
250 km northeast of Los Angeles, 34.9° N, 116.2° W), near the Broadwell Lake playa at an elevation of 400 m. Local mean annual air temperature is estimated to be 22°C (hyperthermic soil temperature regime), and mean annual precipitation is estimated to be 100 mm (aridic soil moisture regime) based on elevation and plant community composition (Miles and Goudie, 1997). The slope at the study site is 2%, with a southeast aspect. Mesozoic granitic and Miocene volcanic rocks (Rogers, 1975) are the dominant sources for the sandy alluvium from which soils at the study site formed. Although creosote-bush is the dominant plant species, the local plant community also includes burrobrush [Ambrosia dumosa (Gray) Payne], brittlebush (Encelia farinosa Gray), and annual grasses and forbs.
Field Sampling
We selected eight creosote-bush shrubs, approximately equally spaced around the perimeter of a circular area about 75 m in diameter. These shrubs were numbered 1 through 8 and were the primary focus of our sampling. We also obtained samples from eight additional shrubs around this perimeter for a few of the analyses described below. Three-meter-long transects were established north and east of each shrub. The maximum shrub diameter and the radius of each shrub along the north transect axis were measured. Shrub maximum diameters ranged from 1.7 to 4.0 m (average of 3.0 m). On each transect, we located sampling sites immediately adjacent to the main stem (0 m, which we refer to as "base" samples) and at horizontal distances of 1, 2, and 3 m from the base. For 14 of the 16 shrubs, the sampling sites at 1 m from the base were under the shrub canopy. Field measurements tested the effect of soil water content on soil respiration. We pushed 5-cm-high by 10-cm-diameter polyvinyl chloride (PVC) rings about 1 cm into the surface soil at the transect sampling sites. On the north transects of four of the shrubs (Shrubs 2, 4, 6, and 8), we added 100 mL of water to each ring to simulate a rainfall event of about 1.3 cm, which we estimated would wet the soils to a depth of about 15 cm. The north transects of Shrubs 1, 3, 5, and 7 were not watered. After about 24 h, soil respiration was measured in the watered and unwatered rings using an EGM-1 infrared gas analyzer (PP Systems, Haverhill, MA), with the following operating conditions: chamber diameter, 100 mm; chamber volume, 1170 cm3; headspace sampled every 8 s, for 120 s final reading; fan speed, 300 to 350 cm3 min–1.
We attempted to measure hydraulic conductivity of the soil at several locations with a constant-head permeameter, but hydraulic conductivity was too rapid to obtain reliable results. Based on this field experience and soil texture, we estimated the hydraulic conductivity to be in the 10 to 100 µm s–1 range (Soil Survey Division Staff, 1993).
A soil pit was excavated near the base of one of the shrubs. The pedon was described and sampled by horizon for particle size analysis. In addition, soil samples were collected from the (approximately) upper 2 cm of soil within each PVC ring at each of the unwatered transect sampling sites (0, 1, 2, and 3 m), and immediately outside each PVC ring at the watered transect sampling sites for lab analyses. We assumed that microbial communities would be most concentrated very close to the soil surface in this hot, dry climate, so we focused our attention on the upper 2 cm for the majority of our analyses. The upper few centimeters of soils are also those most likely to be affected by localized erosion and deposition (e.g., Kieft et al., 1998). These samples, and samples from six additional shrubs with north and east transects (14 shrubs total), were transported in plastic bags to the lab at ambient temperature for analysis of field-moist water content and pH, carbonate content, total C and N content, and 15N content of air-dried soil. For DNA, PLFA, and soluble C and N analyses, an additional set of samples was collected from the upper 2 cm of soils at the unwatered north transect sampling sites (0, 1, 2, and 3 m) of Shrubs 1, 3, 5, and 7. These samples were transported on ice, stored frozen (–20°C), and processed frozen (not air dried).
Lab Analyses
Particle size distribution of air-dried soil material was measured on the pedon samples by the pipette method (Soil Survey Staff, 1996). Unwatered samples from all 14 shrubs were analyzed as follows. Gravimetric water content of the field-moist samples was calculated by mass difference after 105°C oven drying overnight. Total C and N and 15N in air-dry soil were measured using an Integra-CN integrated combustion, purification, and measurement system (Europa Scientific, Crewe, UK). Carbonate was measured in samples from north and east transects of Shrubs 1, 3, 5, 7, 9, and 13 by reacting 1 g of soil with 12 mL of 12 M HCl in a closed jar (similar to the method of Soil Survey Staff [1996]). The resulting CO2 gas was quantified using a Horiba infrared gas analyzer (Southeastern Automation Group, Knoxville, TN). Carbonate concentrations were calculated by comparing the evolved CO2 to standard curves prepared from CO2 gas standards of known concentration, as well as CO2 generated by combining analytical-grade CaCO3 with acid. Soil pH of all samples from Shrubs 1, 3, 5, and 7 was measured in 1:1 soil/water mixtures with a glass electrode (Soil Survey Staff, 1996).
Soluble C and N were measured in the frozen samples from north transects of Shrubs 1, 3, 5, and 7 (unwatered). Three replicate soil extracts were prepared from each sample by standard KCl extraction (Keeney and Nelson, 1982), and frozen (–20°C) for subsequent analysis. The KCl extracts were diluted 1:4 with ultrapure (18 M
) water before analysis for dissolved organic C (DOC) using a Phoenix 8000 UV-persulfate C analyzer (Tekmar-Dohrmann, Cincinnati, OH). Dissolved NH4+ and NO3– in the extracts were measured colorimetrically using a QuickChem 8000 automated ion analyzer (Lachat Instruments, Milwaukee, WI).
Frozen samples from the north transects of Shrubs 1, 3, 5, and 7 (unwatered) were analyzed for PLFA. Triplicate 5-g subsamples were extracted with a one-phase solvent extractant using a modification of the method of Bligh and Dyer (1959). Polar lipids (including phospholipids) were separated from neutral and glycolipids using solid-phase extraction columns (0.50 g Si, Supelco, Bellefonte, PA). The polar lipid fraction was subjected to mild alkaline methanolysis as described previously (Bossio and Scow, 1998), and the resulting fatty acid methyl esters (FAMEs) were extracted with two aliquots of hexane. The hexane was evaporated under N2 at room temperature and FAMEs derived from polar lipids were redissolved in hexane containing an internal standard (19:0 FAME). Samples were analyzed by capillary gas chromatography (HP 6890, Agilent Technologies, Palo Alto, CA) using a 25-m Ultra-2 column (J&W Scientific, Agilent Technologies, Palo Alto, CA) using 1:50 split injections, H2 carrier gas, and a programmed temperature increase from 170 to 260°C at 2°C min–1. Peaks were identified using 33 bacterial FAME standards and MIDI peak identification software (MIDI, Newark, DE).
Fatty acids were described using the nomenclature "number of carbons: number of unsaturations," followed by double-bond locations referenced from the omega (), or aliphatic, end of the molecule. For example, 18:17 denotes an 18-carbon, monoenoic fatty acid with a double bond at carbon 7. Other conventions are: cy (cyclopropyl group); i, a (or br) iso- or anteiso- (branched [or unspecified branching]); 10me methyl- (branched at C10 [from carboxylic end]); and c (cis orientation at double bond).
Soil DNA was measured in frozen samples from the north transects of Shrubs 3 and 7 (unwatered). Triplicate 10-g samples were extracted using buffered sodium dodecyl sulfate, proteinase K, freeze–thaw cycles, bead-beating, chloroform–water separation, and precipitation in ethanol (Zhou et al., 1996). Extracts were purified to generate 400 µL of crude extract. The DNA concentrations were determined from a 1:5 dilution of 100 µL of crude extract, using a Lambda 10 UV/Vis spectrophotometer (PerkinElmer Corp., Norwalk, CT).
Statistical Analyses
One-way analysis of variance was used to identify significant variation in measured nutrients, isotopes, biomass, and other variables with distance from shrubs. Post-hoc, pairwise comparisons were performed using Tukey's HSD test (Zar, 1999), with significance declared at P < 0.05. Simple linear correlation was used to consider relationships between pairs of variables without assuming dependence of one variable on another (Zar, 1999). All analyses were performed using JMP IN version 3.2.6 software (SAS Institute, Cary, NC).
Fatty acid data were analyzed using two statistical ordination methods designed to explore the structure of large multivariate data sets. Ordination methods have the common goal of arranging sample points in space such that points that fall close together have similar values for variables. The first ordination method, principal components analysis (PCA), is analogous to multiple linear regression and is widely used to summarize multivariate data sets of many types (Brereton, 1990; Digby and Kempton, 1987). Fatty acid data for PCAs were converted to mole percentages to eliminate the effect of differences in total PLFA abundance (microbial biomass) among sample locations. The PCA ordinations were calculated using correlation (rather than covariance) matrices (Jackson, 1993, 1997). These PCA ordinations were unconstrained in the sense that no information about potential environmental driving forces (explanatory variables) was included in the calculations. Only relationships among samples and variables (fatty acids) were considered. The constrained, or direct gradient analysis, form of PCA is called redundancy analysis and shows the relationship of samples and variables (fatty acids) to measured environmental gradients (Rao, 1964), in this case the environmental gradient from shrub canopies to spaces between shrubs. The significance of measured environmental gradients can be tested explicitly using a Monte Carlo simulation. The Monte Carlo test returns a P value associated with the effect of the environmental variable on the PLFA composition of the samples. A good explanation of the Monte Carlo test in the context of ordination can be found in Verdonschot and Ter Braak (1994). Ordination analyses were performed using Canoco software (Microcomputer Power, Ithaca, NY).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
15N values also varied significantly with distance away from shrubs, increasing from 4 at 0 m to 7 at 3 m (Table 1). Following wetting, soil respiration (measured as the flux of CO2–C) tended to be higher at the base than between shrubs, but measured rates were highly variable (CVs in Table 1), probably due to contributions from roots of irregularly distributed grasses and forbs. The means for CO2–C with distance (Table 1) were correlated with those for C (R2 = 0.91) and PLFA (R2 = 0.92), even though the individual data points for each shrub were not. The data did not resolve mechanisms controlling higher respiration with proximity to the shrub base. Variability in the respiration measurements suggests that more intensive sampling would be required to firmly establish the relationship between respiration and microbial biomass (activity), and between respiration and microbial community (activity of specific organisms).
|
The PLFA and DNA analyses (Table 1) indicate that base soils supported significantly larger microbial populations than soils between shrubs. The estimates of microbial biomass from PLFA and DNA correlated well (R2 = 0.61). The PLFA content was strongly correlated with total C (R2 = 0.80) and total N (R2 = 0.78). Although the relationship between DNA concentration and distance was not highly replicated (n = 2), the DNA content was also strongly correlated with total C (R2 = 0.86) and total N (R2 = 0.89).
The PLFA results (Table 2) indicate that distinct microbial populations occurred in base samples compared with other locations (Fig. 1 ), even though most of the 1-m sampling sites were under the shrub canopy. The two axes shown in Fig. 1a represented 48% of the total variability in the data set. Distance from the shrub was a significant explanatory variable, suggesting that base soil microbial communities were distinct from those between shrubs. Base and interspace samples both contained actinomycetes (high G+C, gram positive bacteria), indicated by fatty acids that are methyl-branched at the 10th C atom ("10me" prefix in Fig. 1b; Kroppenstedt, 1985; Kampfer and Kroppenstedt, 1996). This was not surprising, as actinomycetes possess a filamentous structure that can give them a competitive advantage in arid settings like the Mojave, where access to nutrients is limited by low water potential (Kinsbursky et al., 1990). When Fig. 1a and 1b are superimposed, the redundancy analysis shows that a subset of actinomycetes (10me18:0) was more prevalent in shrub interspaces, whereas a different subset (10me17:0) was more prevalent at the base of shrubs (i.e., at a0, b0, and c0). Absolute quantities of both 10me17:0 and 10me18:0 were greater at the base, as was total microbial biomass (Tables 1 and 2). Actinomycetes known to synthesize fatty acid 10me18:0 are coryneform bacteria within genera Corynebacterium and Rhodococcus (Kroppenstedt, 1985; Kampfer and Kroppenstedt, 1996). These results suggest that elevated soil nutrients near shrubs supported a larger microbial community, in which some organisms were favored more than others. The presence of actinomycetes at all locations suggests a general microbial community adapted to the arid conditions of the Mojave, with a specific composition that differed between the base of shrubs and the shrub interspaces.
|
|
9). These fatty acids contributed to separation along the first axis (Fig. 1b), which also separated base and non-base samples (Fig. 1a). However, non-plant PLFAs were equally, or more, important for separating samples along the first axis (Fig. 1b), suggesting that plant lipids were not the primary drivers of the variation in PLFA signatures with distance from shrubs.
Proximity to shrubs also resulted in qualitative shifts in the predominant forms of C and N. More than half of the base soil C was organic, whereas soil C was virtually all carbonate between shrubs (Table 1). The N in all samples was mostly organic (>99%, Table 1), but 15N values were lower in base soils (4.1 ± 0.31, Table 1) relative to interspace soils (15N = 6.8 ± 1.34, Table 1), suggesting variation in N sources or, more likely, in transformation and loss of N. Base 15N values were comparable to average values for L. tridentata leaves observed by Shearer et al. (1983) in the Sonoran Desert. If CO3–C is excluded from the calculation of C/N, the ratio tends to decrease from 0 to 3 m (compare with C/N values in Table 1), which suggests that the interspace soil organic matter was more highly decomposed than base soil organic matter. Thus, the enriched interspace 15N values could reflect losses of isotopically light N during decomposition of L. tridentata litter. Alternatively, the C/N ratio may reflect variation in the quality of litter and rates of plant litter input under shrubs compared with spaces between shrubs. Notably, the degree of 15N variation that we observed between base and interspace soils was similar to the variation observed between canopy and intercanopy soils associated with Juniperus osteosperma (Torr.) Little (Utah juniper) in southern Utah (Evans and Ehleringer, 1993).
Soil 15N values may broadly reflect temperature and precipitation, as hotter and drier climates yield increased soil 15N compared with cooler, wetter climates (Amundson and Baisden, 2000; Austin and Vitousek, 1998; Schulze et al., 1991). The N isotope differences (Table 1) may indicate that this correlation holds for the presumably cooler and slightly moister microclimate created by the shrub canopy. At this Mojave Desert site, N concentration and 15N were inversely correlated only in the samples at 3 m (R2 = 0.23, Fig. 2
), and were not correlated at other locations under or between the shrubs. We interpret these results to indicate loss of isotopically light N from interspace soils, thereby enriching 15N, but this interpretation warrants further investigation, in particular with respect to soil processes that might underlie our observations.
|
It has been argued that N and water are limiting to net primary productivity in the Mojave (Chew and Chew, 1965; Sharifi et al., 1988; Lajtha and Whitford, 1989). These limitations would not preclude leaching of N from surface soils during short, intense rain events, or during wetter-than-average years, particularly between shrubs, where biological demand for N is presumably lower than under shrubs. The presence of deep (several meters) NO3 reservoirs beneath some desert soils (Walvoord et al., 2003) indicates that deep N transport can occur in aridic climates, and this process is possible at our sites, especially between shrubs. One interpretation of our results is that N was more tightly cycled under wetter, cooler conditions close to shrubs, where plant uptake and growing microbial populations restricted N loss.
Shrub radius correlated positively with PLFA (R2 = 0.65), soil moisture (R2 = 0.96), total soil N (R2 = 0.65), and NH4+ (R2 = 0.64) in base samples, suggesting a positive feedback in which shrubs enhanced resource islands as they grew (or vice versa), and that the effect was most intensely focused closest to the main stem. We speculate that in resource islands, immobilization of N during relatively wet, productive periods in spring, and mineralization of N during drier times, contributed N for relatively rapid plant uptake as the system transitioned to drought conditions.
Because we did not attempt to control for the relative abundance of annuals in intershrub sampling areas, we could not quantify the extent to which annuals might have contributed to greater variability in respiration and N isotopes with distance from shrubs. Although shrubs may control the location of resource islands (Halvorson et al., 1995), competition with grasses (Holzapfel and Mahall, 1999; Caldwell et al., 1985) and proliferation of roots where nutrients occur (Jackson and Caldwell, 1989; Robinson et al., 1999) could have contributed to tighter N cycling in nutrient patches. Future research to characterize microbial community variation in shrub ecosystems should account for plant competition and root plasticity controlling nutrient availability.
In summary, variation in soil microbial community fingerprints accompanied variation in the distribution of soil respiration and C and N with proximity to Larrea tridentata shrubs in the eastern Mojave. Both PLFA and DNA indicated that microbial biomass was greatest beneath the shrubs, a location also characterized by maximum soil respiration rates. The PLFA analyses showed (i) that both base and interspace soil microbial communities contain actinomycetes, organisms that may hold a competitive advantage in regions of low water potential, and (ii) that the composition of the microbial community at the base of the shrubs and in the space between shrubs was different. Fractionating losses of N may dominate soil N inventories between shrubs, and may result in leaching of isotopically depleted N to deep soil reservoirs. The degree to which resource islands limit soil N loss in shrub ecosystems may be a function of how efficiently microbial populations in those islands capture leachable N during rain events.
|
ACKNOWLEDGMENTS |
|---|
Received for publication August 26, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
15N of soils. Oecologia 94:314–317.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
