| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Soil, Water, & Environmental Science, Univ. of Arizona, Tucson, AZ 85721
b Univ. of California, Davis, CA 95616
* Corresponding author (crasmuss{at}ag.arizona.edu).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: Alp, pyrophosphate-extractable aluminum • AN, andesite • BS, basalt; Feo, oxalate-extractable iron • GR, granite • PE, priming effect • PI(t), priming index with time • PLR, percentage litter respired • PP (or pp), ponderosa pine • RF (or rf), red fir • SRO, short-range order • WF (or wf), white fir
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Priming effects depend on substrate composition and concentration, soil properties, and the interaction of substrate with the microbial community. The addition of litter provides an energy-rich, labile source of C, allowing the decomposer community to increase their investment in exoenzymes to make substrates available (Dalenberg and Jager, 1989; Schimel and Weintraub, 2003; Fontaine et al., 2003). Differences in priming with substrate type and concentration may be a function of microbial community adaptation (Hamer and Marschner, 2002). Griffiths et al. (1998) observed structural changes in the microbial community with the addition of varying amounts of synthetic root exudates to soil and suggested that a substrate concentration threshold may regulate microbial activity. The intensity of microbial activation by substrate addition may also depend on soil properties, such as clay content or the preexisting amount of easily decomposable C and available N in the soil. For example, Bol et al. (2003) found a greater priming effect in clayey soils relative to sandy soils, and Rasmussen et al. (2006) found the soil mineral assemblage to significantly affect soil C mineralization in conifer forest soils. In particular, the latter study found that SRO Al and Fe oxyhydroxide and Al–humus complex content was negatively correlated with soil C mineralization.
Substantial research supports the finding that the soil mineral assemblage is important in controlling soil C dynamics, particularly the presence of SRO materials and Al–humus complexes (Martin and Haider, 1986; Veldkamp, 1994; Torn et al., 1997; Percival et al., 2000). As such, we expect the presence of SRO and Al–humus materials to also have significant impacts on the priming of extant soil C. The SRO materials possess a large reactive surface area (on the order of 800 m–2 g; Harsh et al., 2002) and exhibit variable charge due to an abundance of pH-dependent surface charge sites (Parfitt, 1980), both of which facilitate soil C adsorption and the interaction of negatively charged organic functional groups with the mineral surface. In addition, organics may coordinate with surface Al and Fe groups, forming stable organo–mineral complexes (Tate and Theng, 1980; Yuan et al., 2000). Surface adsorption of soil C promotes particle cementation and the formation of stable micro- and macroaggregate structures (Oades, 1988), further imparting protection of soil C. These properties should promote recalcitrance of soil C and favor the decomposition of added litter, thereby limiting the priming of extant soil C.
Environmental factors, such as temperature, also interact with substrate type and soil properties to affect priming. Temperature not only influences the rate of soil C mineralization, but may also control microbial community size and structure. For example, increasing temperature may cause shifts in microbial community structure, exoenzyme production, and the ability of the microbial community to access recalcitrant soil C pools (Zogg et al., 1997; Andrews et al., 2000; Waldrop and Firestone, 2004). Further, Nicolardot et al. (1994) found a significant interaction between temperature and microbial utilization of various substrate types. The interaction of environmental factors, such as temperature, with substrate quality and the soil mineral assemblage remain inadequately understood and quantified, representing a knowledge gap on control of soil C maintenance.
We examined a range of temperate conifer forests with varying parent material and degree of soil development in the Sierra Nevada of California. The objectives of this study were: (i) to determine the influence of conifer litter addition, temperature, and soil mineral assemblage on the priming of extant forest soil C; and (ii) to determine which of these factors are important for predicting the response of native soil C pools to climate change and litter addition.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Mineralogical Analyses
Briefly, qualitative mineralogical analysis by x-ray diffraction (XRD) was conducted on the clay (<2-µm) fraction of one pedon from each sample site. Oriented clay mounts with standard treatments of Mg saturation, Mg saturation and glycerol solvation, K saturation, and heat treatment of K-saturated samples at 350 and 550°C (Whittig and Allardice, 1986) were analyzed with a Diano XRD 8000 diffractometer (Diano, Woburn, MA), producing Cu-K
radiation. Quantitative mineralogical analysis of bulk soil by selective dissolution procedures was performed using standard methods of acid ammonium oxalate, sodium pyrophosphate, and citrate–dithionite extraction (Soil Survey Staff, 2004). Extractions were not sequential. Oxalate extracts Al, Fe (Feo), and Si from organic complexes and SRO Fe oxyhydroxides (e.g., ferrihydrite) and aluminosilicates (e.g., allophane and imogolite). Pyrophosphate extracts Al (Alp) bound in organo–metal complexes. Sodium dithionite extracts Fe (Fed) from organic complexes, and secondary forms of Fe oxyhydroxides, both crystalline and noncrystalline (Parfitt and Childs, 1988; Dahlgren, 1994).
Carbon-13-Labeled Litter
We grew seedlings of ponderosa pine in a sand–vermiculite–peat medium in a climate-controlled, Plexiglas chamber to label them with 13C following methods outlined in Horwath et al. (1994). Plants were exposed to 375 mg L–1 13CO2 for one photoperiod once a week throughout the "growing season" (approximately 5 mo in duration). Night respiration was allowed to be reassimilated to increase 13C-CO2 uptake efficiency. Following several cycles of bud formation and full needle expansion, water was withheld to force senescence. After senescence, litter material was air dried and litter quality quantified by proximate analysis at the Natural Resources Research Institute at the University of Minnesota–Duluth, following the methods of Ryan et al. (1990) and McClaugherty et al. (1985). Total C, N, and 13C content of ball-milled litter samples were quantified by high-temperature dry combustion (Carlo-Erba Elemental Analyzer, CE Elantech, Lakewood, NJ) and continuous-flow isotope ratio mass spectrometry (Europa Hydra 20/20, PDZ Europa, Rudheath, UK) at the University of California–Davis Stable Isotope Laboratory.
Incubation Design and Procedures
Surface soils from each forest type–parent material combination were incubated at 5.0, 7.5, and 12.5°C with and without the addition of 13C-labeled PP litter (LT and NO treatments, respectively). The incubation temperatures represent the mean annual soil temperature (MAST) of the RF, WF, and PP forest types, respectively. The suite of temperatures allowed examination of how native soil C responds to litter addition at the native forest-type MAST, as well as the effect of rising temperature on soil C response to litter addition.
Soil material from the surface A horizon (approximately equivalent to a 0–10-cm depth increment) of three pedons from each site was composited for use in the incubations. Four replicate (30 g air dry) samples of composited soil were placed in specimen cups and brought to 55 to 60% of field capacity (Cassel and Nielsen, 1986) with deionized water. The 13C-labeled litter (cut to roughly 1-cm lengths) was added to the samples at a rate of 0.01 kg C kg–1 soil. Litter and soil material were thoroughly mixed with a spatula to incorporate litter into the soil matrix. Soils were then tamped down to a uniform depth and bulk density (30 g of soil tamped to a volume of 40 mL or an approximate bulk density of 0.8 g cm–3) with a glass rod and placed in 1-L Mason jars with septa to allow for headspace sampling. The jars were then placed in the dark in temperature-controlled chambers.
The CO2 concentration of the headspace was measured using an infrared gas analyzer (Qubit CO2 Analyzer, Model S-151, Qubit Systems, Kingston, ON, Canada) at frequent intervals. Headspace samples were initially collected daily when rates of CO2 production were high, and the sample interval extended to once a week by the end of the 90-d incubation, following the decreased rate of CO2 production with time. Starting on Day 5 of the incubation and repeating at 15-d intervals, 12-mL gas headspace samples for 13C isotopic analyses were collected coincident with CO2 concentration sampling. Isotope samples were transferred to vacuum-evacuated tubes and analyzed for 13C-CO2 by isotope ratio mass spectrometry (Europa Geo 20/20, PDZ Europa) at the University of California–Davis Stable Isotope Laboratory. The 13C-CO2 content was reported on a per-mil basis using the standard convention relative to the Pee Dee Belemnite standard (Balesdent and Mariotti, 1996). Sample jars were vented after each CO2 concentration and isotope sampling. Headspace samples from blank jars were used to correct the sample CO2 concentration and the 13C/12C isotope ratio for ambient CO2 concentration. Ambient concentration values for CO2 ranged from 0.03 to 0.05%, and ambient 13C-CO2 isotope values ranged from 8 to 10
.
Statistical Analysis
Calculation of CO2 evolution at each sample period followed Zibilske (1994). The CO2 percentage was converted to a mass of C and normalized to soil and litter C content (C activity basis) (g C kg–1 soil C + litter C) following correction with the blank value. The 13C signature of the headspace CO2 was used to calculate the fraction of CO2 derived from litter (F value) following blank correction:
 | [1] |
SOM is the 13C signature of the soil organic matter, CO2 is the 13C of respired CO2, and litter is the 13C signature of the litter material (Balesdent and Mariotti, 1996). The fraction of CO2–C derived from native soil C pools was then estimated by difference (e.g., soil C = 1 – F). The 13C data were collected less frequently than CO2 concentration data. Therefore, we estimated F at each CO2 concentration sample period by assuming a linear relationship in F between 13CO2 sample points. The use of 13C-labeled litter allowed calculation of the amount of CO2–C derived from existing soil and litter using the F value. Soil C mineralization (rate and amount) could then be compared between LT and NO samples, and the effects of litter addition and temperature on priming of soil C estimated.
We calculated a priming index with time [PI(t)] that compares the effects of litter addition on the rate of native soil C mineralization (Kuzyakov et al., 2000; Hamer and Marschner, 2002). The PI(t) parameter was estimated using soil C mineralization rate data from both LT and NO treatments:
 | [2] |
The overall priming effect (PE) was estimated from the cumulative native soil CO2 production from LT and NO treatments:
 | [3] |
We tested for differences in cumulative soil and litter C mineralization and PE between the factorial combination of parent material (AN, BS, or GR) and forest type (PP, WF, or RF) using a two-way ANOVA and Tukey's HSD post-hoc test (P 
0.05). We also tested for differences in cumulative C mineralization and PE by temperature using a one-way ANOVA. We do not present the factorial combinations of temperature x forest type and temperature x parent material as no significant interactions were found. Multiple linear regression analysis between PE, mineralogical properties (clay content and selective dissolution extracts), and soil C quality properties (C/N and 13C of soil C) at each temperature (5.0, 7.5, and 12.5°C) was used to examine the influence of parent material and soil C on PE. Soil data used in the linear regression analysis were presented previously in Rasmussen et al. (2006). The AN, GR, and BS samples were pooled within each temperature treatment for regression analyses (n = 9). In addition, we pooled data across all treatments (n = 27) for an overall multiple linear regression analysis. Regression and ANOVA analyses were performed with JMP 5.1 (SAS Institute, Cary, NC).
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Proximate analysis data indicated the ponderosa pine litter used in the incubation contained roughly 45% C and had C/N = 21 and lignin/N = 11. The N concentration (2.1%) of the litter was typical of young, 1-yr-old ponderosa pine foliage, e.g., leaf litter values for productive ponderosa pine stands averaged 1.8% (Gomez et al., 2002). The ponderosa pine litter contained acid-soluble (43%) and water-soluble (30%) components, with roughly 23% lignin. The litter proximate analysis suggested a relatively labile C source, as indicated by the low C/N and lignin/N and the moderately sized water-soluble component (Scott and Binkley, 1997). Isotopic analysis of the ponderosa pine litter indicated substantial 13C incorporation in the plant material, with an average 13C value of 717 (±8). The substantial enrichment of the litter relative to extant soil C (ranging from –29 to –23) provided a ready means to discriminate between the relative sources of respired C by quantifying the 13C-CO2 of headspace samples.
Cumulative Carbon Mineralization
An ANOVA of CO2 production parameters (cumulative CO2, percentage of CO2–C derived from soil and litter, and PLR) revealed several significant patterns and differences between parent material, forest type, and temperature (Table 1, Fig. 1). Cumulative CO2 (soil + litter derived C) was significantly greater in GR soils, followed by BS and AN. The percentage of total CO2–C derived from soil and litter, as revealed by ANOVA and the isotopic signature of respired CO2 with time (Fig. 1), also varied significantly among parent materials. As presented in Fig. 1, greater 13C enrichment of respired CO2 indicates greater litter C mineralization; litter 13C enrichment averaged 717, so 13C-CO2 values approaching 700 indicate near-complete dominance of respired CO2 by litter resources. Of the total mineralized C, the GR and BS soils mineralized significantly greater soil C than AN soils (soil CO2–C was only 17% of the total in AN soils relative to 34 and 33% in the BS and GR parent materials, respectively). Low total CO2 production and the majority of CO2–C derived from litter indicate relatively resistant soil C in AN soils compared with the other parent materials. Previous studies have also demonstrated soil C protection in andesite-derived soils, particularly Andisols; Al–humus complexation and adsorption to SRO mineral surfaces were proposed as stabilization mechanisms (Torn et al., 1997; Parfitt et al., 2002; Rasmussen et al., 2005). The relatively high rate of litter addition (0.01 kg C kg–1 soil) may account for some of the overall dominance of respired CO2 by litter sources. The percentage of added litter respired as CO2 (PLR), did not show any variation among parent materials. A lack of significant difference in PLR between parent materials demonstrates that variation in total CO2 production was dominated by variation in soil C utilization. The specifics of soil and litter C mineralization within each parent material, forest type, and temperature are further discussed below.
|
|
Ponderosa Pine Forest Type
In the PP forest type, each parent material demonstrated substantial increases in cumulative CO2 production with an increase in temperature, particularly between 7.5 and 12.5°C (Table 2). Cumulative CO2 production also varied among parent materials, with the general pattern of GR > BS > AN, similar to the significant differences observed in the overall data set (Table 1). The isotopic signature (Fig. 1) and calculated percentage of CO2–C derived from soil and litter sources also varied with temperature and among parent materials (Table 2). Generally, the proportion of respired CO2 derived from soil C sources increased with temperature, with greater soil C utilization at 12.5°C relative to 7.5 and 5.0°C in all of the parent materials. The PLR also increased with temperature in all parent materials, but there was little difference in PLR between parent materials at each temperature (Table 2). Increased soil C utilization and PLR with temperature demonstrated a temperature effect on both litter and soil C mineralization, with increased temperature facilitating greater mineralization of both C sources. The GRpp soils demonstrated the greatest proportion of soil-derived CO2 at all temperature treatments, whereas ANpp demonstrated the least soil C utilization. Variation in soil C use by parent material was in agreement with overall data set trends (Table 1) showing significantly less soil C as a percentage of the total CO2 respired in AN soils. Variation in soil C mineralization by parent material (i.e., GRpp soils contain the most labile soil C and ANpp soils the most resistant) suggests significant control of soil C availability, and we propose that soil mineralogy plays a dominant role in this control.
|
Red Fir Forest Type
The general pattern of RF forest type CO2 production followed the WF and PP forest types, where GR > BS = AN (Table 2). But, unlike the PP and WF forest types, there was no difference in CO2 production between the BSrf and ANrf soils. The BSrf soils contain substantially more soil C than GRrf and ANrf soils (161 vs. 53 and 97 g kg–1, respectively), such that on a C-activity basis, BSrf CO2 production was minimized. The percentage of total CO2 derived from soil C sources was substantially greater in the BSrf forest type at each temperature, however (Fig. 1, Table 2). For example, nearly 60% of the BSrf CO2 was derived from soil C sources at 12.5°C, compared with 30% in GRrf and 20% in ANrf soils, respectively. A greater percentage of CO2 derived from soil C sources was probably a function of the large soil C stocks providing a proportionately larger pool of soil C to mineralize.
Priming Response
Overall ANOVA analyses indicated significant variation in the PE by parent material (Table 1). In particular, AN forest types exhibited four to six times less priming effect than GR and BS soils, respectively. The variation in PE by parent material indicates that soil properties, such as clay content or the soil mineral assemblage, may be responsible for the observed differences in soil C mineralization. The PE did not exhibit significant differences by forest type or temperature, further highlighting the importance of parent material as an underlying control of priming.
Ponderosa Pine Forest Type
The overall priming effect (PE) in the PP forest type indicated that addition of ponderosa pine litter increased soil C mineralization in all parent materials (Table 2), but that the magnitude of the increase varied with both temperature and parent material. The ANpp and GRpp PE values tended to decrease with an increase in temperature, suggesting a greater effect of litter addition on soil C mineralization at cold temperatures (5.0 and 7.5°C). At colder temperatures, addition of the labile, energy-rich litter appeared to facilitate decomposition of soil C sources that were otherwise temperature limited. The temperature limitation of soil C mineralization in NO samples may derive from temperature-controlled variation in the active microbial community or exocellular enzyme production (Waldrop and Firestone, 2004). We hypothesize that litter addition provided a labile energy source that enhanced production of exoenzymes, further facilitating mineralization of soil C.
Whereas both GRpp and ANpp soils demonstrated a decrease in PE with increasing temperature, BSpp soils demonstrated a positive interaction between temperature and litter addition, particularly at 12.5°C where priming induced a 73% increase in soil C mineralization. This was in conjunction with a near threefold increase in cumulative soil C mineralization (8.7 to 25.4 g C kg–1 soil C) in BSpp compared with more moderate increases in cumulative soil C mineralization (one to two times) in GRpp and ANpp, respectively (Table 2). The highly variable response of PE and soil C mineralization to temperature indicated significant variation among parent materials within the same forest type. The soil mineral assemblage may exert significant control over soil C recalcitrance that feeds back to the ability of the microbial population to access soil C pools. Differences in PE and soil C mineralization by parent material may also be attributed to variation in the microbial community or inactivation of enzymes by the mineral soil matrix (Burns, 1986).
The priming index with time, PI(t), which is a measure of the relative increase or decrease in soil C mineralization rates with time following litter addition, demonstrated substantial temporal variation for each parent material and temperature treatment in the PP forest type (Fig. 2). Positive priming effects, which indicate an increased rate of soil C mineralization with litter addition, were most predominant at the beginning of the incubation, with a steady decrease of PI(t) values to near or below zero with time (except for BSpp 12.5°C, which is discussed below). The length of time required for PI(t) to decrease toward zero increased with decreasing temperature, e.g., from 5 d at 12.5°C, to 15 d at 7.5°C, to >20 d at 5.0°C (Fig. 2). An initial rapid priming response was in agreement with previous studies (Kuzyakov et al., 2000; Hamer and Marschner, 2002; Bol et al., 2003) and suggests priming due to co-metabolic decomposition of soil C resources. It has been hypothesized that the addition of labile, energy-rich litter facilitates increased exoenzyme production and utilization of soil C resources for microbial metabolism (Fontaine et al., 2003). The use of a labile energy source for production of enzymes to utilize more complex soil C resources represents the aforementioned co-metabolic decomposition. The observed greater time for decrease in PI(t) at colder temperatures may be a function of reduced microbial activity or biomass production.
|
White Fir Forest Type
The WF forest type priming effects varied greatly among parent materials. Both GRwf and BSwf soils exhibited substantial increases in soil C mineralization with ponderosa pine litter addition (Table 2). In particular, BSwf soil C mineralization nearly doubled, with a PE of 92%, following addition of ponderosa pine litter at 12.5°C. The ANwf soil C mineralization was little affected by ponderosa pine litter addition, with a negative PE at 5.0°C and only a 4% increase in soil C mineralization at 12.5°C. The ANwf soils also produced substantially less CO2 than the other WF parent materials, with the majority of CO2 derived from litter sources (Table 2). Reduced PE in ANwf soils suggests protected soil C or an inhibition of microbial growth and enzyme activity. Short-range-order Al oxyhydroxides may significantly reduce enzyme activity through physical adsorption processes (Saggar et al., 1996; Miltner and Zech, 1998), while Al–humus complexes and exchangeable Al may have toxic effects on the soil microbial community (Illmer et al., 2003). A mineralogy-induced recalcitrance of soil C and inhibition of soil C mineralization was supported by the fact that PLR varied little between parent materials (Table 1 and 2), suggesting that the microbial population in ANwf soils was capable of decomposing the ponderosa pine litter, but could not access existing soil C resources for metabolic processes.
Litter addition impacts on the rate of soil C mineralization, PI(t), also varied substantially between WF parent materials (Fig. 2). All parent materials exhibited the greatest priming effect at the beginning of the incubation, followed by a steady decline in PI(t) with time. The BSwf soils tended to show a secondary increase in soil C mineralization rates midway through the incubation that may be a function of microbial biomass turnover and nutrient release as discussed above. The ANwf PI(t) values peaked during initial incubation stages, followed by negative PI(t) in the middle of the incubation, and a subsequent increase in PI(t) at the end of the incubation in each of the temperature treatments (Fig. 2). The dual peak suggests that initial soil C mineralization rates were stimulated by litter addition and then declined as labile soil C was exhausted. Recovery of soil C mineralization rates at the end of each temperature treatment implies secondary labile energy and nutrient sources that facilitate metabolism of less labile soil C resources. It should be noted that the time necessary for the ANwf PI(t) to decline to negative values increased at colder temperatures. The change in response time of soil C mineralization rates to litter addition, coupled with little to no change in the total soil C mineralized in ANwf soils (Table 2) suggests a fixed pool of labile soil C at each temperature, with the main effect of litter addition being a more rapid depletion of this labile pool. Negative PI(t) values suggest that once this labile pool was exhausted, further soil C mineralization and mineralization rates were limited.
The effect of litter addition on the ANwf soil C mineralization rate was particularly evident at colder temperatures, with much greater PI(t) at 5.0°C than at 12.5°C (Fig. 2). The addition of energy-rich litter may have facilitated the decomposition of labile soil C at cold temperatures that was otherwise temperature limited. It is interesting to note that cumulative ANwf soil C mineralization (g C kg–1 soil C) was similar between temperature treatments at the transition point between positive and negative PI(t) values, e.g., 1.1, 1.3, and 1.3 g C kg–1 soil C for the 12.5, 7.5, and 5.0°C treatments, respectively. Assuming that this transition point represents the exhaustion of labile soil C resources, a similar quantity of soil CO2–C production at this crossover point suggests a fixed labile soil C pool, with the turnover rate of this pool dependent on temperature. In NO treatments, exoenzyme production may be limited by temperature at 7.5 and 5.0°C, but with ponderosa pine litter addition, a labile energy source is provided, enabling greater exoenzyme production relative to NO treatments, and hence a greater rate of soil C mineralization. In contrast, GRwf cumulative soil C mineralization at the PI(t) positive-to-negative transition point increased from 9.5 to 14.7 g C kg–1 soil C with increasing temperature. Coupled with substantial increases in GRwf PE with increased temperature, these data suggest that litter addition facilitated mineralization of a larger pool of GRwf soil C, and the magnitude of this increase was greater at warmer temperatures.
Red Fir Forest Type
The RF soils demonstrated substantial variation in priming effects with temperature and parent material (Table 2). Both BSrf and GRrf PE values increased with temperature, indicating a positive interaction between temperature and litter addition on soil C mineralization. The BSrf soils exhibited the greatest priming effect, particularly at 12.5°C, where PE was nearly 90%. In conjunction with significant priming, BSrf soils mineralized nearly 60% of their total CO2 production from soil C sources (Table 2), suggesting that BSrf soil C stocks were more labile than the added ponderosa pine litter. In contrast, ANrf PE values decreased substantially from 26 to 3% at 7.5 to 12.5°C, respectively. The ANrf data suggest that addition of litter primed a temperature-limited soil C pool at 5.0 and 7.5°C, but that this soil C was available at 12.5°C regardless of litter addition.
The effects of litter addition on RF soil C mineralization rates varied significantly between parent materials (Fig. 2). The BSrf soils exhibited a secondary peak in PI(t) in all temperature treatments, with the secondary peak most pronounced in the 12.5°C treatment. The ANrf and GRrf soils exhibited an initial increase in C mineralization with litter addition that slowly tapered off with time. The tail of PI(t) was generally more pronounced in the 5.0 and 7.5°C treatments, suggesting that cold temperatures delayed complete mineralization of the primed soil C. The ANrf cumulative soil C mineralization at the transition point between positive and negative PI(t) was similar in all temperature treatments (2.0, 1.9, and 1.7 g C kg–1 soil C at 12.5, 7.5, and 5.0°C, respectively), suggesting a fixed labile C pool with a temperature-dependent turnover rate. These data and the decreased PE with temperature indicate that the effect of litter addition on priming was limited at warmer temperatures. Reduced priming impact at warmer temperatures may be a function of soil C recalcitrance and fixed mineralizable pool size, or inhibition of exoenzyme activity by the mineral soil matrix, both of which are controlled to a certain extent by the soil mineral assemblage.
In contrast to ANrf, GRrf cumulative soil C mineralization at the PI(t) positive–negative transition increased substantially with temperature (5.1 to 8.5 g C kg–1 soil C at 5.0 and 12.5°C, respectively). A greater labile pool size with temperature suggests limited soil C recalcitrance. The GRrf soils were dominated by crystalline minerals (such as kaolinite, gibbsite, and hydroxy-interlayered vermiculite), with limited surface area and charged sites for soil C adsorption, both of which limit the C protective capacity of the GRrf soil mineral assemblage. The variation in priming dynamics between parent materials strongly suggests that the soil mineral assemblage was a significant control of the soil C mineralization response to litter addition.
Regression Analysis of Priming Effects
Multiple linear regression analysis of PE using all the treatment data pooled into one data set indicated significant control of PE by both soil C quality (soil C/N and 13C enrichment) and soil mineral assemblage (Feo and Alp) variables (Table 3). The PE demonstrated a significant negative correlation with soil C/N and a significant positive relationship with soil 13C enrichment. Soil C/N ratios ranged between 20 and 30, whereas soil C 13C values ranged from "depleted" (–29) to "enriched" (–23). The negative correlation with C/N suggests that litter quality, in the form of N availability, limits priming and soil C mineralization in forest soils with high C/N ratios. Relatively high soil C/N appeared to promote utilization of litter resources rather than soil C resources. Similarly, organic compounds such as microbial metabolites, carbohydrates, and sugars tend to exhibit an enriched 13C signature, whereas phenolic and lignin components generally possess a depleted 13C signature (Benner et al., 1987). Hence, greater PE in soils with an enriched 13C signature suggests a relatively labile C source that may be related to soil C quality and dominance of soil C components by carbohydrates and microbial byproducts. Variation in soil 13C may be due to differences in the 13C of litter inputs or variation in soil C quality resulting from variability in water and N availability.
|
Opposite to expectations, Alp, a measure of Al–humus complexation, demonstrated a significant positive correlation with PE. Generally, Al–humus complexation is considered to stabilize soil C against decomposition (Percival et al., 2000). The positive correlation between PE and Alp suggests that the addition of a labile, energy-rich C source enabled the decomposition of an otherwise stable soil C pool in the form of Al–humus complexes. We hypothesize that litter addition facilitated the production of exoenyzymes with the ability to decompose Al–humus complexes or promoted growth of microorganisms with the ability to utilize these compounds. Rasmussen et al. (2006) found an increasingly negative impact of Al–humus content on soil C mineralization at colder temperatures. The greater impact of Al on C mineralization at colder temperatures suggested a temperature limitation to the exoenzyme production needed to decompose the Al–humus complexes. In this experiment, litter addition may have overcome this limitation and may explain the pattern in ANrf PE described above. It should also be noted that at cold temperatures, soil mineral variables explained the majority of variation in PE, while at warmer temperatures, soil C quality variables became more important (Table 3). Regression analyses suggest that interactions between temperature and soil chemical and physical properties control PE.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Priming effects indicated varying interactions between temperature and litter addition on soil C mineralization. In particular, AN soils tended to show a decreased effect of litter addition on soil C mineralization with temperature that may be related to mineral-moderated soil C recalcitrance or adsorption of exoenymes. The PI(t) data indicate that AN soils tend to have a fixed labile pool size, with a significant impact of litter addition on the turnover rate of that fixed pool. In contrast, GR and BS soils tended to show a positive interaction between temperature and litter addition, with an increase in both mineralizable pool size and mineralization rate with litter addition and increasing temperature. The PI(t) data suggested that the majority of priming effects were short term, possibly resulting from stimulation of microbial growth, increased biomass, or exoenzyme production facilitated by the addition of a labile, energy-rich litter resource. Furthermore, increases in PI(t) at later stages of the incubation and secondary PI(t) peaks suggested stimulation of various microbial populations that utilize microbial metabolites as an energy source, thereby enabling continued co-metabolic decomposition of soil C material.
While this study demonstrated clear effects of parent material and soil C quality on soil and litter C mineralization and priming effects, further study is needed to characterize possible variation in microbial community composition and exoenzyme production by parent material within a particular forest type. It is possible that the soil mineral assemblage selects for specific microbial communities that exhibit varying response to litter addition and temperature. Experiments characterizing the interaction of substrate type, soil mineralogy, temperature, and active microbial community composition would greatly improve our understanding of soil C dynamics and priming effects.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication October 31, 2006.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. L. Parfitt Allophane and imogolite: role in soil biogeochemical processes Clay Minerals, March 1, 2009; 44(1): 135 - 155. [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 | |||
