| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Environmental Studies, Univ. of California, Santa Cruz, CA 95064
b Dep. of Environmental and Resource Sciences, Univ. of Nevada, Reno, NV 89557
c Dep. of Nematology, Univ. of California, Davis, CA 95616
* Corresponding author (wxcheng{at}ucsc.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: DAP, days after planting • DI-H2O, deionized water • PVC, polyvinyl chloride • SOM, soil organic matter
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Roots have been found to have both stimulatory and inhibitory effects on SOM decomposition. Laboratory experiments have shown that when 14C-labeled plant material was decomposed in soils planted with maize (Zea mays L.), spring wheat, or barley (Hordeum vulgare L.), 14CO2 release from the soil was reduced compared with bare soil (Reid and Goss, 1982, 1983; Sparling et al., 1982). In contrast, a stimulatory effect has been reported in other laboratory experiments using 14C-labeled litter (Helal and Sauerbeck, 1986; Cheng and Coleman, 1990). Furthermore, other research has shown that SOM decomposition is dependent on the length of exposure to living roots. In a 2-yr study, roots suppressed the decomposition of newly incorporated 14C-labeled plant material during the first 200 d but stimulated the mineralization of the 14C-labeled materials in the soil during the latter stage when compared with bare soil (Sallih and Bottner, 1988). Recently we have found that different combinations of plant species and soils resulted in different rhizosphere effects on SOM decomposition (Fu and Cheng, 2002). Several studies have also suggested that soil mineral nutrition is an important modifier of rhizosphere effects (Merckx et al., 1987; Liljeroth et al., 1994; Ehrenfeld et al., 1997); the direction of rhizosphere effects may change depending on the level of the soil mineral nutrients. However, most studies have focused on the occurrence of rhizosphere effects, with little attention being paid to the multiple factors that may influence it.
Studies of rhizosphere processes have been restricted by the limitations of existing methods. Separating soil C from current plant-derived C has been accomplished by using isotope techniques. Total rhizosphere respiration is often defined as the sum of root respiration and rhizomicrobial respiration of substrates derived from roots. Total rhizosphere respiration and original soil C decomposition have been quantified by either continuous 14C labeling (Barber and Martin, 1976; Whipps and Lynch, 1983; Liljeroth et al., 1994) or pulse labeling (Cheng et al., 1993). However, both of these 14C-labeling methods have limitations. Continuous 14C-labeling requires special facilities and often requires transplanting of seedlings which may have considerable unlabeled food reserves, therefore requiring some time for all plant parts to become evenly labeled. Because pulse labeling does not uniformly label all plant C, total plant-derived C cannot be separated from soil-derived C in a pulse-labeling experiment. Because of the safety issue when radioactive materials are used, most 14C-labeling experiments have been of short duration. A 13C natural tracer method for measuring total rhizosphere respiration and soil organic C decomposition has recently been developed (Cheng, 1996; Qian et al., 1997; Rochette and Flanagan, 1997). This natural tracer method eliminates some of the major limitations of earlier labeling methods, and makes possible for systematic investigations of rhizosphere effects on SOM decomposition.
The main objective of this study was to investigate simultaneously three potentially important factors: plant species, plant phenology, and fertilization, that may significantly control the magnitude and the direction of the rhizosphere effect on SOM decomposition. Research questions were: (i) What are key interactions between different factors that control the direction and magnitude of rhizosphere effects? (ii) What is the relative significance of each controlling factor when all factors are studied together in one experiment? (iii) How does the rhizosphere effect change as controlling factors change through time?
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
13C values, between plants with C3 vs. C4 photosynthesis pathways. C3 plants such as wheat or soybeans are depleted in 13C and have a mean 13C of -27
, while C4 plants such as maize have a mean 13C of -12 (Smith and Epstein, 1971). The 13C values of belowground C roughly reflect the original source of plant C.
If C3 plants are grown in a C4–derived soil, the C entering the soil via roots of the C3 plants has a different 13C value than the 13C value of the soil. The following equation can be used to partition soil-derived C4–C from plant-derived C3–C (Cheng, 1996):
 | [1] |
t is the 13C value of Ct, 3 is the 13C value of the C3 plant C, and 4 is the 13C value of the C4–soil C.
Results from several studies have indicated that this method worked as expected (Cheng and Johnson, 1998; Cheng et al., 2000; Fu and Cheng, 2002). No isotopic fractionation was found between total root C and rhizosphere CO2–C since the 13C values of both C sources obtained from a C-free inoculated sand treatment were not statistically different from each other (Cheng, 1996). Analysis of 13C abundance of all samples was performed using a PDZ Europa (Cheshire, UK) Hydra 20-20 continuous flow isotope ratio mass spectrometer coupled with an automated C and N combustion analyzer at University of California-Davis Stable Isotope Facility.
Soil Used
Surface layer soil (
0–30 cm) was obtained from a tallgrass prairie field at the Konza Prairie Long-Term Ecological Research site. Vegetation at this site has been dominated by C4 grasses. The 13C value of the total soil organic C is -14.2. The soil is a clay loam, and classified as a Mollisol. The soil contains 2.3% organic C and 0.2% N, and has a pH of 7.6. The soil was sieved (<2 mm), homogenized, and air-dried before use.
Experimental Design
Spring wheat and soybeans were grown in C4 soils with three levels of fertilization (F0, F1, F2) and four sequential destructive sampling dates. The four sequential destructive sampling dates represented vegetative (35 d after planting or DAP), flowering (DAP = 68), grain-filling (DAP = 89), and maturity (DAP = 110) growth stages for wheat, and early-vegetative, late-vegetative, flowering, and grain-filling growth stages for soybeans, respectively. A complete factorial randomized design was employed, with treatments replicated four times. A no-plant control was also included. Fertilization treatments were done by applying NH4NO3 for N, and KH2PO4 for P and K in a solution at the time of initial watering to air-dried soils. Deionized water (DI-H2O) was used for fertilization level one (F0, zero NPK added). For the second fertilization level (F1), 500 mg of N, 50 mg of P, and 63 mg of K were added per container (equivalent to 288 kg N, 29 kg P, and 36 kg K ha-1). For the third fertilization level (F2), 1000 mg of N, 100 mg of P, and 126 mg of K were added per container (equivalent to 566 kg N, 57 kg P, and 71 kg K ha-1).
Plant Growth
A plant growth apparatus and a closed-circulation CO2 trapping system are illustrated in Fig. 1. Polyvinyl chloride containers (15 cm i.d., 40 cm long) were used to grow plants. Each container was the basic unit for all treatments and measurements, and was instrumented for measuring soil respiration. A closed-circulation system was used for trapping total CO2 from the soil-root system. The potential problem of minor air leakage was minimized because of its closed-circulation design and the use of a Mylar balloon for air pressure buffering.
Each container was packed with 7000 g of air-dried soil to a predetermined bulk density (1.2 kg L-1), which is equivalent to 6250 g of oven-dried (at 105°C) soil in each container. Sixteen seeds were planted into each container for the wheat treatment, and four seeds per container for the soybean treatment. For even distribution of applied fertilizers, fertilization treatments were applied at the first watering of the air-dried soil in each container. Nutrient solutions at two concentrations were prepared according to the fertilization treatments specified in the experimental design section and the amount of water to be applied to each container. Air-dried soils in each container were watered using either DI-H2O (F0, zero fertilization), or nutrient solutions (F1 and F2) to 80% field water holding capacity. All containers were weighed on a top-loading balance after watering. All seeded and watered containers were randomly distributed in an area in a greenhouse. After emergence, plants were thinned to one plant per container for soybeans, and 12 plants per container for wheat. Water loss in each container was determined gravimetrically using a top-loading balance with a sensitivity of ±1 g. Each container was watered with DI-H2O daily according to the amount of water loss to avoid a soil moisture difference between the planted treatments and the no-plant treatments. To prevent possible occurrence of anaerobic conditions, all containers were aerated for 30 min. every 6 h throughout the duration of the experiment by connecting an aquarium air pump to each container. A digital timer was used to control the aeration system. The experiment started on 5 Mar. 1999 and lasted until 3 July 1999 for a total period of 120 d. Maximum air temperature in the greenhouse ranged from 25 to 33°C. Minimum air temperature ranged from 17 to 24°C. Relative humidity ranged from 15 to 90%.
Measurements
The method of Cheng (1996) was used to measure SOM decomposition and total rhizosphere respiration. Briefly, before each destructive harvesting, four containers from each treatment were sealed at the base of the plant with low melting point Paraffin (m.p. 42°C) to separate the aboveground atmosphere from the belowground atmosphere. The integrity of the seal was verified by submerging the PVC container in water and checking for gas leaks. The Paraffin seal did not damage the plant seedlings, and was strong enough to withstand routine fluctuation of air pressure. The CO2 was trapped for 48 h in a column containing 30 mL of 4 M NaOH solution mixed with acid-washed sand (Fig. 1). The CO2 trapping efficiency of this system was nearly 100% as checked by an infrared gas analyzer (Model LI-6262, LI-COR, Lincoln, NE). Thus, preferential sorption of 13CO2 vs. 12CO2 was eliminated. After CO2 trapping, the contents (sand and NaOH) of each column were transferred into a plastic bottle with a sealed cap by rinsing with DI-H2O. The bottle was capped and shaken vigorously. After settling overnight, an aliquot of the NaOH solution in each plastic bottle was analyzed for total inorganic C using a TOC Analyzer. Another aliquot of the trapping solution was mixed with excess SrCl2, and the 13C of the precipitate (SrCO3) was analyzed by mass spectrometry (Harris et al., 1997). The amounts of soil-derived CO2 and plant-derived CO2 were separated using the 13C method described above. Soil-derived CO2 was used as a measure of original SOM decomposition. Plant-derived belowground CO2 was used as the measure of total rhizosphere respiration. Analytical and experimental controls were employed during the 13C analysis. Null CO2 trapping units (blanks) without soil-roots were included to correct for contamination from carbonate in the NaOH stock solution and from sample handling. The average amount of inorganic C trapped by four replicate null units was used as a measure of the amount of contamination in all samples. The highest amount of contaminant C among all samples was <5% of the total C. The effect of the contaminant C on the 13C value of each sample was corrected using the following equation:
 |
j is the 13C value from a sample after correction, t is the 13C value of a sample before correction, c is the 13C value of the contaminant C (-8; the value of the CO2-3 in NaOH stock and from the air); Ct is the total amount of C in the sample solution including contaminant C; Cc is the amount of C in the blank control as contamination.
After each destructive sampling, plant shoots were cut at the soil surface and oven-dried at 65°C for 48 h before weighing. Roots were hand-picked, washed, and oven-dried at 65°C before weighing. After grinding, the 13C value of each shoot or root sample was analyzed for isotopic composition. Total soil inorganic C content was analyzed using an infrared gas analyzer. After grinding and mixing in a SPEX ball mill, 500 mg of oven-dried soil sample was put into a 125-mL flask. The flask was then connected to a CO2–free carrier gas system at a constant flow rate of 500 mL min-1. Carbonates in each soil sample were purged for 30 min by injecting 20 mL of acid solution (1 M H2SO4 with 5% FeSO4 as antioxidant) under magnetic stirring. The CO2 concentration of the outflow gas was recorded digitally at 1-s intervals during the entire purging period. The amount of total inorganic C in each sample was calculated using the total area under the curve and known CaCO3 standards.
Statistical Analysis
All statistical analyses of data were performed using Statistix 7 software (Analytical Software, Tallahassee, FL). This experiment consisted of three factors (plant species, levels of fertilization, and growth stages) with four replicates. Growth stages were not considered repeated measures because they were represented by sequential harvests of random replicates. Analysis of variance (ANOVA) of all data sets was performed using the general linear model with all possible interacting terms. Mean comparisons were performed using Fisher's LSD method.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
Natural 13C Analysis
The average 13C value was -27.59 for roots of wheat or soybean. No significant difference in 13C values was detected among C3 plant materials. The 13C value of -27.59 was used to represent the 13C value of plant-derived C in this experiment, because there was no detectable isotopic discrimination associated with rhizosphere processes (Cheng, 1996).
As expected, the 13C value of CO2 evolved from the no-plant control remained in the range of -13 to -14 during the whole experiment, indicating that the origin of the C was primarily from C4 sources (Fig. 3)
. The 13C value of CO2 evolved from the wheat treatment was low (mostly from the C3 source) at the first sampling, increased almost linearly as sampling stages progressed to the end of the experiment, indicating that the portion of wheat-derived C in the total belowground CO2 decreased. On the contrary, the 13C value of CO2 evolved from the soybean treatment was high (near the value from the C4 soil only) at the first sampling and decreased linearly during the course of the experiment, indicating that the portion of soybean-derived C in the total belowground CO2 increased as soybean plants grew and approached maturity. Analysis of variance indicated that both the plant factor (no-plant, wheat, or soybean) and the time of sampling significantly influenced the 13C values of belowground CO2, but the overall effect of fertilization was not statistically significant (Table 3).
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The C4 soil used in this experiment had a pH value of 7.6 and contained on average 0.57 mg of inorganic C g-1 of soil (Fig. 4). No statistically significant rhizosphere effect on soil inorganic C content was found at the end of the experiment. This result indicated that the enhanced soil-derived CO2 effluxes from planted treatments mostly came from rhizosphere-stimulated SOM decomposition instead of soil carbonates. The cumulative amount of soil C loss in the soybean treatment averaged across all fertilization levels was 1.44 mg C g-1 soil, which is 0.90 mg C g-1 soil higher than the control (0.54 mg C g-1 soil). This extra C loss caused by the soybean rhizosphere effect was even greater than the total soil inorganic C. The cumulative soil C loss in the wheat treatment averaged across the three fertilization levels was 1.07 mg C g-1 soil, or 0.52 mg C g-1 soil higher than the no-plant control, or nearly twice as much as the total inorganic C.
Because of the rhizosphere effect, cumulative original soil C loss averaged across all fertilization levels during the entire growing season was 164% (or 0.90 mg C g-1 soil) and 96% (or 0.52 mg C g-1 soil) more than that in the no-plant control for soybean and wheat treatments respectively (Fig. 6). The amount of soil C loss due to soybean rhizosphere effect in another experiment (Fu and Cheng, 2002) using the same C4 soil was 70% higher than the no-plant control, much lower than the level found in our current study. Differences in environmental conditions and soybean varieties between the two experiments might have caused these apparently dissimilar levels of rhizosphere effects. The greenhouse used for this study generally had higher air temperatures, higher light intensities (full sunlight vs. lime shaded), and much higher variation in temperature and humidity. Different soybean varieties were chosen because of different locations where the two experiments were conducted. The real causes of these dissimilar levels of rhizosphere effects on SOM decomposition warrant further studies.
The amount of extra soil C loss due to rhizosphere effects during the entire growing season is approximately the same amount of C in standing root biomass for soybean, and 71% of root biomass C for wheat. If all C inputs and outputs are considered, the planted soil systems should have net C gains, as C inputs from root standing biomass, aboveground litter, and rhizodeposition more than compensated for the total C loss during the growing season. However, directly applying our current results to a field situation is probably unwise because soil disturbance during preparation and greenhouse operation may have exacerbated the rhizosphere effect. Averaging across the three levels of fertilization, the amount of C loss in the form of soil-derived CO2 during the entire growing season represents 2.4, 4.6, and 6.3% of the total initial soil C for the no-plant control, wheat, and soybean treatments, respectively. These values are in a reasonable range compared with the value of 6% found in the field study using the 13C method (Rochette et al., 1999). Planting density may also have influenced the level of rhizosphere effects in the greenhouse experiment. The level of rhizosphere effects on the decomposition rate of 14C-labeled litter ranges from 71% reduction (Sparling et al., 1982) to as high as 350% increase (Helal and Sauerbeck, 1987) above the no-plant control. Litter decomposition is often different from decomposition of mineral-associated SOM because of differences in substrate quality, decay dynamics, and environmental conditions. Therefore, the results obtained from experiments using labeled litter should be interpreted as rhizosphere effects on litter decomposition instead of on SOM decomposition. Negative rhizosphere effect on 14C-labeled litter decomposition were reported earlier (Reid and Goss, 1982, 1983; Sparling et al., 1982). They also reported that a significant amount of 14C-labeled C was taken up by plant roots, which might have contributed to the negative effect on the decomposition rate. However, this root uptake hypothesis was largely dismissed (Sallih and Bottner, 1988; Cheng and Coleman, 1990). The high level of rhizosphere priming effects on the decomposition rate of 14C-labeled litter reported by Helal and Sauerbeck (1987) was estimated by budgeting 14C-labeled materials remaining in the planted soil zones, in the soil zones away from roots, or in the unplanted control rather than by direct measurement of soil C loss as 12CO2. Because of the inherent high variability of SOM, the budgeting approach was widely acknowledged to have low sensitivity and low accuracy. Because our present study employed a natural 13C tracer method, the rhizosphere effect on SOM decomposition was measured directly in the form of soil-derived CO2. The above-mentioned limitations associated with experiments using either 14C-labeled litter or C budgeting were avoided. The level of sensitivity of the natural 13C tracer method in this study was higher than that reported early by Cheng (1996). Standard errors of 13C values ranged from 0.07 to 1.13, with a mean of 0.33. The difference in 13C values between the two sources (plant-derived and soil-derived) was 13.6. The accuracy of the natural 13C tracer method used in this study ranged from 74 to 98% with a mean of 92%, using a 95% confidence level (t0.025 = 3.182, n = 4 - 1). For the level of rhizosphere effects found in this study, the accuracy of the natural 13C method was adequate beyond any reasonable doubt.
Using continuous 14C labeling technique in a relatively short experiment (50 d) with wheat as the test plant, Liljeroth et al. (1994) reported a rhizosphere priming effect of 200%. Helal and Sauerbeck (1984)( 1986) reported much higher rhizosphere priming values of 336 and 432%. However, these higher values were indirectly derived from soil C budgets with large measurement errors caused by high variability in soil C data. Nevertheless, the consistently lower amounts of soil C remaining in the soil sections with roots or closer to the rooting zone than sections away from the rooting zone indicate priming effects by roots in their studies. These continuous labeling studies require the use of a special labeling facility which only permits short experimental duration and limited plant growth space. Because of these limitations, the duration of the two experiments mentioned above was relatively short, roughly corresponding to the first or the second sampling of this present study. However, as shown here, plant phenology (or sampling times) was the most important variable in controlling the amount of rhizosphere-primed soil C loss. The magnitude of rhizosphere effects changed significantly through time. Moreover, the amount of rhizodeposition and root activities differs at different plant growth stages (Cheng et al., 1990; Martin and Merckx, 1992), resulting in different rhizosphere effects at different phenological stages.
The rhizosphere effect of soybeans is significantly different from wheat. This finding is further confirmed by another study using four plant species (Fu and Cheng, 2002). The stronger rhizosphere effect of soybeans may have been caused by the higher substrate quality of soybean rhizodeposition because the N concentrations of soybean tissues are generally higher than tissues of nonleguminous plants. The N-rich compounds usually produce stronger priming effects than compounds of low or no N (Dalenberg and Jager, 1989). This clearly demonstrates that plant species is an important factor controlling SOM decomposition through rhizosphere processes. The effects of plant species on the rhizosphere effects on SOM decomposition have rarely been considered.
In general, fertilization did not significantly affect SOM decomposition in this experiment, with or without a rhizosphere. This is a surprise because several studies have suggested that soil mineral nutrition is an important modifier of rhizosphere effects (Merckx et al., 1987; Jackson et al., 1989; Schimel et al., 1989; Liljeroth et al., 1994; Ehrenfeld et al., 1997). Fertilization reduced rhizosphere effect on SOM decomposition as shown by some studies using soils from agricultural fields and 14C-labeling (Merckx et al., 1987; Lekkerkerk et al., 1990; Liljeroth et al., 1994). This negative impact of fertilization on the rhizosphere effect has been put forward as a hypothesis of preferential substrate utilization (Cheng, 1999). This hypothesis states that, when mineral nutrient supply is abundant, soil microorganisms prefer labile root-derived C to soil-derived C, resulting in a decreased rhizosphere effect on SOM decomposition. If mineral nutrients are in short supply, soil microorganisms prefer nutrient-rich SOM to root-derived C, resulting in increased SOM decomposition. However, some other studies (Jackson et al., 1989; Schimel et al., 1989; Ehrenfeld et al., 1997) seem to suggest that the presence of the rhizosphere reduces SOM decomposition in nutrient-poor soils, possibly because of the competition for mineral nutrients between roots and rhizosphere microbes. The extra root-derived C input in the rhizosphere may decrease SOM decomposition because the increased root uptake of mineral nutrients in the rhizosphere may retard microbial growth under mineral nutrient-poor conditions. The result of our experiment supports neither of these hypotheses because fertilization did not significantly modify the rhizosphere effects on SOM decomposition. This apparent controversy may be caused by the different soils, plant species, environmental controls, and experimental methods used in the various studies.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication September 1, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. E. Kolb, K. J. Fermanich, and M. E. Dornbush Effect of Charcoal Quantity on Microbial Biomass and Activity in Temperate Soils Soil Sci. Soc. Am. J., May 13, 2009; 73(4): 1173 - 1181. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Lehman, S. L. Osborne, and K. A. Rosentrater No Evidence That Bacillus thuringiensis Genes and Their Products Influence the Susceptibility of Corn Residue to Decomposition Agron. J., November 7, 2008; 100(6): 1687 - 1693. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Shennan Biotic interactions, ecological knowledge and agriculture Phil Trans R Soc B, February 27, 2008; 363(1492): 717 - 739. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M.-F. Johnson, R. R. Allmaras, and D. C. Reicosky Estimating Source Carbon from Crop Residues, Roots and Rhizodeposits Using the National Grain-Yield Database Agron. J., April 11, 2006; 98(3): 622 - 636. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. A. Dijkstra, S. E. Hobbie, and P. B. Reich Soil Processes Affected by Sixteen Grassland Species Grown under Different Environmental Conditions Soil Sci. Soc. Am. J., March 29, 2006; 70(3): 770 - 777. [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 | |||
