SSSAJ 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 Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (27)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Schjønning, P.
Right arrow Articles by Christensen, B. T.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Schjønning, P.
Right arrow Articles by Christensen, B. T.
Agricola
Right arrow Articles by Schjønning, P.
Right arrow Articles by Christensen, B. T.
Related Collections
Right arrow Soil Organic Matter
Right arrow Soil Models
Right arrow Soil Physics
Soil Science Society of America Journal 67:156-165 (2003)
© 2003 Soil Science Society of America

DIVISION S-3—SOIL BIOLOGY & BIOCHEMISTRY

Linking Soil Microbial Activity to Water- and Air-Phase Contents and Diffusivities

Per Schjønning*,a, Ingrid K. Thomsena, Per Moldrupb and Bent T. Christensena

a Danish Institute of Agricultural Sciences, Department of Crop Physiology and Soil Science, Research Centre Foulum, P.O. Box 50, DK-8830 Tjele, Denmark
b Aalborg University, Department of Environmental Engineering, Sohngaardsholmsvej 57, DK-9000 Aalborg, Denmark

* Corresponding author (per.schjonning{at}agrsci.dk)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Quantification of in situ soil microbial activity is indispensable to improve manipulation of nutrient turnover in soil and optimize crop nutrient supply. We sampled 100-cm3 cores of undisturbed arable soil at three locations along a naturally occurring clay gradient (L1: 11% clay; L3: 22% clay; L5: 34% clay). The cores were drained to seven different matric potentials in the range -15 to -1500 hPa and gas diffusion determined prior to a 4-wk incubation at 20°C in the dark. For all soils the net nitrification increased with water content to a maximum (L1, 12.1 µg NO3–N g-1 soil; L3, 10.3 µg NO3–N g-1 soil; and L5, 8.2 µg NO3–N g-1 soil) and then decreased with further increase in water content. The water content at maximum nitrification was 0.26, 0.37, and 0.42 m3 m-3, respectively. Calculations of water-filled pore space (WFPS) did not normalize soil type differences in optima for microbial activity. The matric potential at peak net nitrification was -140, -170, and -430 hPa, respectively. No single correlation between CO2 evolution and soil water content existed across soil types. The relative solute diffusivity estimated by recently developed models offered a better description of CO2 evolution. The relative gas diffusivity was a better predictor of the increase in net nitrification than was the soil air content. A conceptual model balancing the effects of solute and gas diffusivity indicated that the relative trend in the observed optima of water contents across soil types was as expected. We advocate the use of the conceptual model including soil type dependent expressions for solute and gas diffusivity in future studies of aerobic microbial activity.

Abbreviations: BBC, Buckingham-Burdine-Campbell • BET, Brunauer-Emmett-Teller • CEC, cation-exchange capacity • ODR, oxygen diffusion rate • SA, surface area • WFPS, water-filled pore space


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE EFFECT OF MOISTURE on soil microbial activity is substantial and subject to intense research, and opinions differ on the main mechanisms behind the moisture effects observed in various experimental set-ups. Orchard and Cook (1983) stated that incubation studies should be performed at well-defined water potentials to allow comparisons across differently textured soils since the water retention characteristic (i.e., the relation between the potential and the volume of water in soil) is strongly influenced by soil texture. However, they were intrigued to find that the respiration rate in their soil was directly proportional to gravimetric soil water content. We found a similar response across a range of texturally differing soils drained to specific water potentials (Thomsen et al., 1999). Net N mineralization rates have also been shown to correlate linearly with the soil water content (Stanford and Epstein, 1974; Myers et al., 1982). Griffin (1981) concluded that the direct effects of soil water potential upon microbial activity are important only at low potentials. This was confirmed by Stark and Firestone (1995), who found that cell dehydration became rate limiting for nitrification only at potentials more negative than -6000 hPa. At higher potentials, substrate limitation was reported as the rate-limiting factor. Thus, whereas soil matric potential may be considered an important measure of water availability, soil water content may appear to be a more useful parameter when water availability is not limiting for microbial activity (Skopp et al., 1990).

Several studies have attempted to normalize the effect of water content on soil microbial activity by calculating the WFPS (e.g., Linn and Doran, 1984; Scott et al., 1996; Franzluebbers, 1999). This could facilitate experimental studies as well as model prediction because the WFPS may be estimated in any soil by simple measurements of bulk density and gravimetric water content. A conceptual model balancing the limiting effects of substrate and oxygen diffusion was suggested by Skopp et al. (1990) and several studies support the basic assumptions used in this model (e.g., Stark and Firestone, 1995; Zak et al., 1999). The Skopp model applied the WFPS term as the expression of soil moisture. However, it remains to be shown, whether an index like WFPS provides a better description of the influence of water than the water content in absolute terms.

Many incubation studies have employed homogenized and sieved (often air-dried) soil samples in which aggregates were broken down before incubation (e.g., Myers et al., 1982). Previously we have shown that the physical characteristics of soil exposed to such treatments were quite different from those of undisturbed field soil even after a 17-mo period of structure regeneration (Schjønning et al., 1999). Only a few investigations have employed undisturbed field sampled soil cores in studies of soil microbial activity (e.g., Cabrera and Kissel, 1988; Van Gestel et al., 1992; Stenger et al., 1995).

This study examines the effects of the soil-water regime on microbial activity in undisturbed soil cores of different texture. The range of water contents was chosen to include the expected optimum for aerobic activity. The aim was to evaluate the conceptual model of Skopp et al. (1990) and identify the terms by which the soil water regime regulates the processes.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soils
Replicate 100-cm3 cores (diam. 61.0 mm, height 34.2 mm) of undisturbed soil were sampled at three locations along a naturally occurring texture gradient from an arable field at Lerbjerg in Denmark (56°22'N lat., 9°59'E long.). The field had been cropped to winter wheat (Triticum aestivum L. ssp. vulgare) for several years with pig slurry being applied in the spring. Soil sampling took place in April 2000 prior to slurry application. The cores were collected from the 4- to 8-cm soil depth in stainless steel cylinders at three locations labeled L1, L3, and L5. These correspond to NA1/RE1, NA3/RE3, and NA5/RE5, respectively, in Schjønning et al. (1999) and Thomsen et al. (1999). Each sampling location was approximately 1 m2 and their relative position in the field is illustrated in Fig. 1 . By this sampling strategy, differences in soil management history and mineralogical composition were eliminated (Schjønning et al., 1999).



View larger version (41K):
[in this window]
[in a new window]
  		Fig. 1. Map of the sampling location at Lerbjerg with kriged contours of soil clay content. The area shown is located approximately 20 m from the border of a field of approximately 20 ha. The positions of the sampling sites L1, L3, and L5 are shown together with those of the L2, L4, and L6 sites included in previous studies (e.g., Schjønning et al., 1999; Thomsen et al., 1999).

 
Soil Physical Measurements
The soil cores were weighed, placed on tension tables, and saturated with water from beneath. For estimation of preincubation NO3–N content, two ceramic discs were placed on the soil surface at full water saturation (Strong et al., 1997; Thomsen and Schjønning, 2003). The ceramic discs were allowed to equilibrate with the soil water for 4 h. The soil cores were randomly sorted into seven groups, each with 15 replicates. The groups were drained to obtain matric potentials ({Psi}) of -15, -30, -60, -100, -200, -500, or -1500 hPa, respectively. This was achieved by the use of tension tables (sand or ceramics) for all potentials, except -1500 hPa, where we used pressure plates.

After reaching the specified matric potential, the two ceramic discs were removed and the height of each soil core within the metal cylinder was determined with a purpose-built caliper. Three replicate measurements were performed on each core. The true soil volume was calculated using the measured core height rather than the height of the metal cylinder. Porosity was calculated for each individual core, combining bulk density with average soil particle density. Soil air content at a given matric potential was calculated as the difference between total pore volume and the volume of water retained at that potential.

The soil cores were analyzed for air diffusivity according to Taylor (1949) and as described by Schjønning (1985). Soil gas diffusion was determined at 20°C with O2 as the experimental gas. Our calculations for similar, undisturbed soil cores indicated that we could ignore the O2 consumption in the cores during diffusion measurements. The O2 diffusion coefficient in soil, DS,g, is presented relative to that in free air, D0,g, that is, the relative gas diffusivity equals DS,g/D0,g. Air permeability was measured by a steady-state method (Iversen et al., 2001; Ball and Schjønning, 2002). Before measuring the diffusivity and permeability, we pressed the soil gently at the very edge of the metal ring to minimize the risk of air leaks at the soil–metal ring interface.

Incubation
Following determinations of the gas diffusivity and permeability, the soil core was gently pushed halfway out of the cylinder and sliced horizontally into halves. This was done to allow the samples to serve as a reference for faeces-amended samples treated similarly. The results of this concurrent study are reported elsewhere. Each cylinder was placed on a metal mesh (mesh size of 4 by 4 mm) in sealed 2-L jars and kept in the dark at 20°C. A beaker with water (10 mL) was placed in each jar to minimize desiccation.

Six replicate cores per soil and matric potential were incubated for 28 d. Evolved CO2–C was absorbed during the incubation in 15 mL of 1 M NaOH. The NaOH was renewed on Day 14. Loss of water from the soil cores was examined on Day 14 by weighing. If any water had been lost, a similar amount of water was supplied with a mist sprayer. At the end of the incubation, the soil was removed from the cylinder, carefully mixed and 30 g of soil was immediately extracted in 100 mL of 1 M KCl. The remaining soil was dried at 80°C for 48 h.

Relationship between Inorganic Nitrogen Determined Destructively and by Ceramic Discs
Three additional soil cores from each soil and matric potential were supplied with two ceramic discs at full water saturation before drainage. The ceramic discs were allowed to equilibrate for 4 h whereafter soils were drained to -15, -30, -60, -100, -200, -500, and -1500 hPa. When the matric potential was reached, the ceramic discs were removed for extraction. The soil from each cylinder was mixed immediately after removal of the discs and subsamples were extracted for nitrate content as described above. The remaining soil was dried at 80°C for determination of dry matter content.

Analyses
After removal from the soil, the ceramic discs were weighed and shaken end-over-end in 10 mL of 1 M KCl for 4 h (90 rpm). The discs were removed from the KCl, dried (80°C) and reweighed. The content of mineral particles >2 mm was determined in all soil samples by wet-sieving, and reported results are based on soil <2 mm. The nitrate content in the KCl extracts was determined on a Technicon Autoanalyzer II (Bran+Luebbe GmbH, Norderstedt, Germany). Total CO2–C was determined by HCl titration of excess NaOH after precipitation of CO2 with BaCl2.

An equation was produced to predict the soil nitrate content at Day 0 from measurements of nitrate in the ceramic discs. Net nitrification during the incubation was calculated by subtracting the estimated nitrate content at Day 0 from the nitrate content measured at the end of incubation (see Thomsen and Schjønning [2002] for further details).

Statistical treatment of the data was performed using the SAS statistical analysis system (SAS Institute, 1988).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Basic Soil Characteristics
The content of clay and silt in the soils is reflected in their cation-exchange capacity (CEC) and the surface area (SA) of mineral particles (Table 1). The L5 soil with higher values of clay, CEC, and SA had a lower bulk density and thus a higher porosity than the L3 and especially the L1 soil. The different soil structure properties derived from these basic characteristics are reflected in the different pore-size distributions of the soils (Fig. 2) . The water potential at field capacity for Danish soils is approximately -100 hPa (Madsen, 1976). The arrows in Fig. 2 thus indicate that the soils had not reached field capacity (approximately 30-µm pore diameter) at sampling in early spring. Figure 2 also shows that the loamy sand soil (L1) can be expected to exhibit an air-filled pore volume of approximately 17% (v/v) at field capacity, whereas the corresponding values for the L3 and L5 soils are approximately 6% and approximately 5%. When the L5 soil was drained to a matric potential of -1500 hPa (approximately 2-µm pore diameter) it still retained 39% (v/v) of water, while the L1 soil held only 18%. The L5 soil cores were nearly impermeable to convective gas transport up to a matric potential of -200 hPa. Even at this potential, it had an extremely poor permeability as compared with the L3 and especially the L1 soil (Table 1). Clearly the substantial differences in the soil water characteristic influence the conditions for microbial activity in these soils.


View this table:
[in this window]
[in a new window]
  		Table 1. Basic characteristics of the soils in investigation.

 


View larger version (51K):
[in this window]
[in a new window]
  		Fig. 2. Pore-size distribution derived from the soil water characteristics assuming d ~ 3000/|{Psi}|, where d is the tube-equivalent pore diameter in micrometers, and {Psi} is the matric potential in hectopascals. {theta}sampl indicates the water regime at sampling.

 
Carbon Dioxide Evolution and Net Nitrification
Net nitrification increased with increasing soil water content up to an optimum and then decreased (Fig. 3 , upper figures). Because of the variability in the soil water characteristic associated with undisturbed soil cores, the matric potentials do not follow exactly the expected order of succession for the L3 and L5 soils. The matric potential assigned to each data point is given in Fig. 3 (lower figures). A third-order polynomial was fitted to the data to estimate the optimum water content and the corresponding maximum of net nitrification (Fig. 3, upper figures). For the L1 soil, low values of net nitrification were obtained at -60 and -100 hPa and these were excluded in the analysis. Despite the variability in the soil-water characteristic of the L3 and L5 soils, a relatively close correlation between water content and net nitrification was observed on both sides of the optima. For the L3 and L5 soils, net nitrification at the upper and lower water potentials were significantly different from those at the intermediate potentials. For the L1 soil, only the net nitrification measured at -1500 hPa differed significantly from those at the other potentials. The optimum water content for net nitrification was 0.26, 0.37, and 0.42 m3 m-3 for the L1, L3, and L5 soils, respectively. The corresponding WFPS were obtained by relating the water content to the porosity of the given soil. The matric potentials were interpolated from the water-retention curve (water content plotted against matric potential).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 3. Net nitrification (upper figures) and CO2 evolution (lower figures) as averaged for each combination of soil and matric potential. Bars represent standard error. The values below the CO2 data points for the L3 and L5 soils indicate the numeric matric potentials assigned to the average results shown in the figure. For the L1 soil, the order of succession of data points along the water content axis was as expected from the matric potentials.

 
Analyses of variance for the net nitrification were based on the water potential closest to the peak net nitrification for each soil. The data measured at -200 hPa were taken as a best approximation for the L1 and L3 soils, while the data measured at -500 hPa were used for the L5 soil (Fig. 3, upper figures). Thus the L5 soil displayed a significantly lower net nitrification at optimum water content than the L1 and L3 soils. Although the L1 soil showed the highest net nitrification, it did not differ significantly from that of the L3 soil.

The CO2 evolution increased significantly with increasing water content and then reached a plateau with no significant differences among water potentials (Fig. 3, lower figures). In accordance with de Jong and Schappert (1972), the variability among replicate samples increased when soils became more wet. The break point in the curve coincides with the optimum water content for net nitrification. The plateau of CO2 evolution was averaged over the samples drained to matric potentials on the wet side of the breaking point. The L5 soil had a significantly higher plateau of CO2 evolution than the L1 and L3 soils.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Methodological Aspects
Linn and Doran (1984) and Doran et al. (1988) identified universal optima of water contents for peak CO2 evolution in soils repacked in beakers before incubation, whereas Scott et al. (1996) did not find an optimum water content when using soil placed in open-ended cylinders. In this study, we applied the latter approach and did not detect a decreased CO2 evolution at high water contents (Fig. 3, lower figures). The surface of a soil core will always support aerobic respiration, and an incubated soil core does not instantly turn from completely aerobic to anaerobic conditions but will gradually become anaerobic from the core center and outwards as the oxygen diffusion rate (ODR) decreases (Bridge and Rixon, 1976). The surface/volume ratio of the soil sample has been found to affect the organic matter turnover (Bhaumik and Clark, 1947). For soil contained in closed beakers, air diffusion will be more restricted except at the upper surface exposed to the air phase. Irrespective of methodology used so far, laboratory incubation cannot fully imitate the conditions of soils in situ. In the field, a given soil volume will be surrounded by soil that most often possesses similar conditions for air and water movement and there will be no specific sample boundary. Recalling these methodological reservations, we attempt to examine in more detail the parameters that regulate soil microbial activity under undisturbed conditions.

Substrate Limited Soil Microbial Activity
The WFPS for optimum net nitrification was 0.63, 0.83, and 0.82 in the L1, L3, and L5 soils, respectively (Fig. 3, upper figures). For the two more clayey soils (L3 and L5), this index of water content thus was a better general expression of the optimal water regime for aerobic microbial activity than the water content in absolute units (L3, 0.367 m3 m-3; L5, 0.418 m3 m-3). However, the optimum WFPS (0.82–0.83) is much higher than reported by other researchers (e.g., Franzluebbers, 1999) and appears not to be ‘universal’ across soil types as the sandy L1 soil revealed a much lower optimum. Stanford and Epstein (1974) found the highest nitrification rates at 80 to 90% WFPS, while Franzluebbers (1999) reported an average optimal WFPS of 42% for a range of soil types.

The evolution of CO2 increased up to a water content similar to that of the optimum for net nitrification (Fig. 3). Also Scott et al. (1996) failed to identify a distinct optimum for C mineralization rates in terms of WFPS. A number of studies argued that the WFPS may be used as a ‘unifying’ parameter for description of aerobic microbial activity in terms of CO2 evolution (e.g., Linn and Doran, 1984; Franzluebbers, 1999). Although they argued that 60% WFPS could be regarded as a universal optimum for aerobic microbial activity, the data of Doran et al. (1988)(1990) included clayey soils with optimum respiration rates at quite higher WFPS's. We conclude that the WFPS index may be less suited to normalize soil type differences in the C and N mineralization in undisturbed soils, and that less empirical approaches based on conceptual models for the microbial activity should be favored.

Collis-George (1959) listed spatial constraints as one of four abiotic factors influencing the activity of microorganisms. Also Grant et al. (1993) mentioned ‘space’ as an important regulator of microbial activity. In a previous study with the L1, L3, and L5 soils, we found that CO2 evolution from native soil organic matter was linearly related to the volumetric water content (Thomsen et al., 1999). Figure 3 (lower figures) similarly indicates an increase in CO2 evolution with water content for each of the three soils. Regarding the soil water volume as an expression of ‘space’ this is in support of the hypothesis raised above. However, a concept of space as a regulator of microbial activity does not clarify the mechanisms involved in the regulation. Grant et al. (1993) claimed that the size of the microbial biomass was able to account for the C mineralization observed in a number of studies. At reduced water levels, soil microbial activity increased with increasing water content (Fig. 3, lower figures), whereas microbial biomass did not (Ingrid K. Thomsen, personal communication, 2002). Our results thus suggest that substrate diffusion rather than space per se controls microbial activity in the L1, L3, and L5 soils at low water contents (e.g., Skopp et al., 1990; Zak et al., 1999). A quantitative analysis can be based on the CO2 evolution data to the left (dry) side of the optima detected in Fig. 3. Recent studies of solute diffusivity in soils have facilitated the prediction of transport by diffusion of solutes in soil water (Olesen et al., 1999, 2001). Olesen et al. (2001) found that the solute-independent diffusivity was well described by the model

[1]
where DS,l is the diffusion coefficient of a given solute in soil, D0,l is the diffusion coefficient of the same solute in free water, {theta} is volumetric water content, and {theta}th is a threshold water content at which solute diffusion is effectively zero, likely because of discontinuous water films at low soil water content. Moldrup et al. (2001) showed that {theta}th may be estimated from the N2–BET surface area of soil minerals

[2]
where SAvol is surface area on a volumetric basis calculated from soil bulk density and N2–BET soil surface area on a gravimetric basis (Table 1).

Figure 4 shows CO2 evolution as related to the volumetric water content of individual soil samples for all potentials to the left (dry) side of the optima found in Fig. 3. The three soils had significantly different relations between evolved CO2 and the water content, which may therefore not be used directly to describe the CO2 evolution from these differently textured soils.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4. Carbon dioxide evolution for individual soil cores drained to the potentials to the dry side of the detected optima for net nitrification. The CO2 evolution is related to (a) volumetric water content or (b) the relative solute diffusivity.

 
Hypothesizing that substrate diffusion is the main regulating process for aerobic microbial activity at low water contents, a soil type independent relation should be expected between microbial activity and relative solute diffusivity, DS,l/D0,l. Figure 4b shows the CO2 evolution as related to the relative solute diffusivity calculated from Eq. [1] and [2]. It appears that the L3 and L5 soils rich in clay displayed an identical relationship between CO2 evolution and relative solute diffusivity. A regression of the combined L3 and L5 data predicted an intercept at the CO2–evolution axis not stastistically different from zero (Fig. 4b). This suggests that the conceptual model is consistent with the data, as no diffusion of substrate would be expected to effectively restrain microbial activity. The relative solute diffusivity therefore appears to be a better expression of the conditions for microbial activity in clayey soils than the water content. The poor model prediction for the more sandy L1 soil may indicate that the threshold water content at which diffusion approaches zero should be lower than estimated from the SA of the soil. The deviation may also be because of tortuosity aspects not dealt with in the studies of Olesen et al. (1999)(2001). For example, the substrates (and extracellular enzymes) diffusing to (and from) the surface of microbes mostly will be macromolecules several times larger than the Cl- ion used by Olesen et al. (1999)(2001) in their studies. Despite the results observed for the L1 soil, the use of a solute diffusivity approach may be superior to relating microbial activity directly to water content. It might be claimed that solute diffusivity is not a parameter generally known for different soil types. However, Olesen et al. (2001) showed that the relative solute diffusivity of a soil may be accurately predicted from the water content in combination with bulk density, clay, and silt content.

Aeration Limited Soil Microbial Activity
Substrate diffusion was hypothesized to be the main rate-limiting factor for aerobic microbial activity in dry soil, whereas O2 diffusion may control the activity in wet soil. Only a few studies have combined a quantitative assessment of aeration potentials concurrent with microbial activity (e.g., Groffman and Tiedje, 1991). Most often the microbial response (e.g., CO2–evolution, nitrification, microbial biomass, specific respiratory activity, respiratory quotient) has been related only to soil water in terms of WFPS (e.g., Stanford and Epstein, 1974; Bridge and Rixon, 1976; Linn and Doran, 1984; Scott et al., 1996; Franzluebbers, 1999). We advocate an approach that considers the soil air phase, because aerobic microbial activity relies intimately on diffusion in air (the diffusion rate of O2 in water is 104 times smaller that in air). Figure 5 (upper figures) shows the relationship between net nitrification and air-filled pore space measured for individual soil cores. We would expect an increase in net nitrification with increasing soil air content as O2 is a prerequisite for nitrification, and this trend was found for samples with an air-filled pore volume up to approximately 0.15, 0.10, and 0.08 m3 m-3 for the L1, L3, and L5 soils, respectively. The relationship is less convincing, however, because of the scatter in data, and the soil air content per se appears not to be the sole factor regulating the net nitrification.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 5. Net nitrification for all individual soil cores as related to soil air content (upper figures) or relative gas diffusivity (lower figures). Note the different scales.

 
The air diffusivity apparently is a better bid for the parameter controlling net nitrification (Fig. 5, lower figures). Although we observe some data scatter, there is a distinct lower threshold in net nitrification for the L3 and L5 soils. The thresholds increase with increased air diffusivity and describe lower boundary lines, indicating optima for net nitrification near DS,g/D0,g = 0.030 (uncertain), 0.006, and 0.004 for the L1, L3, and L5 soils, respectively. Stepniewski (1980)(1981) measured ODR and bulk soil air diffusivity for a number of soils at different bulk densities and water contents. It appeared that a sharp decline in ODR occurred at bulk air diffusivities (DS,g/D0,g) in the range approximately 0.005 to 0.02. Our results are in accordance with these findings and do further suggest that clayey soils (the L3 and L5 soils) have lower thresholds than more sandy ones (the L1 soil). On the dry side of the optima (Fig. 5), air diffusivity has no direct causal relation to net nitrification. At these water contents, substrate diffusion is expected to be rate limiting (Fig. 3 and 4).

Up to the levels of air diffusivity optimal for net nitrification, we consider nitrification to be aeration limited. The state of aeration will determine whether dissolved O2 rather than nitrate is the main electron acceptor in microbial metabolism. In this range of air diffusivity, a balance exists between gross nitrification and denitrification. Note that the lower level of net nitrification in each soil type (values to the very left in the figures) declines with increased clay content (structural complexity) (L1: approximately 5; L3: approximately 0; L5: -4 µg NO3–N g-1 soil). Similarly, the peak net nitrification predicted by the lower boundary line at the optimum air diffusivity depends on soil type (L1: approximately 10–12; L3: approximately 9; L5: approximately 6 µg NO3–N g-1 soil). We take this to be effects of soil type characteristics beyond the effects of air diffusivity. However, air diffusivity controls the concentration of oxygen within the soil volume and hence the production of nitrate, and whether the nitrate produced will be denitrified at microsites with low level of O2. This means that O2 and nitrate diffusion in the water phase determines the balance between gross nitrification and denitrification.

Some of the cores displayed values of net nitrification significantly above the lower boundary line. This is considered to be because of diffusional constraints on nitrate in the denitrification process (Myrold and Tiedje, 1985). These well structured soils (especially the L3 and L5 soils) produce large aggregates, where nitrification may prevail in the surface zone of the aggregates and denitrification dominate in their anaerobic centers (Sexstone et al., 1985). To evaluate this causality, we calculated indices of physical pore characteristics from gas diffusivity and permeability measurements (Ball, 1981; Schjønning et al., 2002). However, none of these parameters were able to serve as a covariable and explain the significant data scatter above the lower threshold line (data and analysis not shown).

The Optimum Water Regime for Aerobic Microbial Activity
The Skopp Model—Fit to Data
The conceptual model of Skopp et al. (1990) suggests that aerobic soil microbial activity (P) as limited by substrate and O2 transport may be represented by the relative diffusivity (DS/D0) of the solute (l) and the gas (g)

[3]
where {alpha} and ß are constants assigned to water or gas concentration gradients and tortuosities of the fluid phase. Please note that {alpha} and ß are not used exactly as in Skopp et al. (1990). The symbol ‘min’ stands for ‘take the minimum of the alternatives given in the brackets’. Equation [3] states that the potential microbial activity is the lesser of the activities calculated from either substrate diffusion or O2 diffusion. As stated by Skopp et al. (1990), it is a mathematical expression of Liebig's law of minimum. In the following, P represents the net nitrification.

Solute diffusion may be predicted from Eq. [1] and [2]. As gas diffusivity was measured in the present study, we may use a model fitted to these data to describe the latter part of Eq. [3]. We selected a simple exponential-type diffusivity model since it could accurately fit our measured data

[4]
where M and N are soil type dependent constants (Table 1), {epsilon} is the soil air content, and {Phi} is soil total porosity.

Figure 6 (upper figures) shows solute and gas diffusivity simulated by the combined Eq. [1] and [2] (solute diffusion), and Eq. [4] (gas diffusion). Measured values of gas diffusivity are also shown. The plots therefore represent Eq. [3] when {alpha} = ß = 1. When the two expressions of the P-min function equal, the conditions for aerobic microbial activity are optimal. The predicted optima ({theta}opt) for {alpha} = ß = 1 are 0.216, 0.248, 0.286 for the L1, L3, and L5 soils, respectively. The relative trend in these theoretical predictions is in accordance with the observed (Fig. 3).



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 6. Relative substrate (broken line) and gas (solid and dash line) diffusivity simulated by selected models. The plots at the upper part show the results if assuming equal importance of solute and gas diffusivity ({alpha} = ß = 1). The plots at the lower part show the results if {alpha} is calculated with a fixed ß = 1 and using the observed optima for soil water content. Data points are mean values of measured relative gas diffusivity for each water potential including all fifteen replicate cores. Bars represent standard error. The values at the data points indicate the numeric matric potentials.

 
It is not possible from our data to derive the true values for {alpha} and ß. However, if we use the observed optima, it is possible to calculate the ratio {tau} = {alpha}/ß. As just stated, the relative solute and gas diffusivity in Eq. [3] will equal at the optima and thus, inserting the expressions given in Eq. [1]/[2] and [4],

[5]

The observed optima of {theta} is labeled {theta}opt, and Eq. [5] may be solved for {tau} = {alpha}/ß

[6]

In Fig. 6 (lower figures), the predictions have been reproduced with ß = 1 and {alpha} as calculated from Eq. [6] for all soils ({tau} = {alpha} = 0.336, 0.031, and 0.045 for the L1, L3, and L5 soils, respectively). Notice that the interceptions of the prediction lines for solute and gas diffusivity ({theta}opt) now correspond with the water potentials (data points) at which the optima were observed (Fig. 3). The simulated values of relative gas diffusivity at {theta}opt (L1, ~0.019; L3, ~0.004, and L5, ~0.005) are in accordance with those derived from individual core data in Fig. 5. The higher {tau} in the L1 soil indicates that gas diffusion is more dominating in this soil than in the L3 and L5 soils (Skopp et al., 1990). However, when using net nitrification to represent microbial activity, it has to be recalled that other processes than solute diffusion may influence net nitrification (e.g., immobilization of nitrate).

The Skopp Model—Trends and Perspectives
In the calculations above, we have considered models that provided the best explanations of trends in data. However, other soil type dependent models exist for both solute and gas diffusivity. A combination of such models may yield an impression of which parameters influence the optimum water regime for aerobic microbial activity across soil types.

Olesen et al. (2001) showed that the threshold water content, {theta}th, for solute diffusivity (Eq. [1]) across a range of soil types was well predicted from the Campbell (1974) water-retention parameter, b, by {theta}th = 0.02 b. The b-parameter is derived as the slope of the regression between matric potential and volumetric water content in a log-log plot. In effect, b is an integrating expression of the pore-size distribution. A soil type dependent expression of solute diffusivity may thus be written as

[7]

Moreover, Moldrup et al. (1999) showed that the Campbell b parameter was able to account for soil type differences in gas diffusivity when combined with soil total porosity and soil air content in the so-called Buckingham-Burdine-Campbell (BBC) gas diffusivity model,

[8]

When inserting these soil type dependent expressions (Eq. [7] and [8]) in Eq. [3], the water content at maximum aerobic microbial activity ({theta}opt) becomes related to {Phi}, b, and {tau} by (in analogy with Eq. [6])

[9]

Quantification of the Campbell b parameter requires measurements of soil water content at a range of matric potentials. A soil water characteristic curve will not be readily available in all studies, but Rolston and Moldrup (2002) showed that the Campbell b parameter can be reasonably accurately predicted from the soil clay content using

[10]
where CF is the soil clay fraction (kg kg-1). By combining Eq. [9] and [10], it is possible to quantify the optimum water content ({theta}opt) for aerobic microbial activity. The {theta}opt is found to increase slightly with increased soil clay content (Fig. 7) . The term {tau} represents an indirect influence of soil texture by integrating effects related to the diffusion pathway (effective diffusion path, tortuosity, and concentration gradients, etc. [Skopp et al., 1990]). It appears from Fig. 7a that the soil type difference in {theta}opt primarily is exerted through this parameter. Note that the total porosity (i.e., the bulk density) has a significant direct influence on {theta}opt (Fig. 7a; {Phi} = 0.4 or {Phi} = 0.6 m3 m-3). When expressing the optimum water content relative to the total porosity (i.e., the WFPS; Fig. 7b), the effect of bulk density is nearly absent. However, the general trend of increased optimum water content with increased clay content and the effect of the complex (and soil type dependent) {tau} parameter is still present. We conclude that the WFPS term is effective in normalizing differences in soil bulk density, but WFPS is not particularly well suited in regulating aerobic microbial activity across soil types.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 7. Theoretical optimum water regime for aerobic microbial activity predicted from the Equation: {tau} = {alpha}/ß = {Phi}2 [({Phi} - {theta}opt)/{Phi}](2+3/b)/[1.1 {theta}opt ({theta}opt - 0.02 b)], and given as (a) the soil water content, {theta}, and (b) the water-filled pore space, WFPS. Solid lines indicate {Phi} = 0.4 and broken lines {Phi} = 0.6. Calculations were performed for three levels of the Skopp parameter, {tau}.

 
The wide range in predicted water content for maximum aerobic microbial activity may explain the conflicting results presented in previous reports. More research on the interactions between the diffusivity of solutes and gases is urgently needed, and the complex {tau} parameter should receive more attention. Nevertheless, we consider the approach suggested by Skopp et al. (1990) and further elaborated in the present study of considerable interest for future studies. Microbial respiration, preferably measured as O2 consumption, will be a particularly interesting parameter in such studies.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Our study showed that it is possible to reveal basic mechanisms responsible for soil microbial activity also in highly variable, undisturbed field soil. Future soil incubation studies intended to simulate field conditions should consider more closely ways to control the air exchange at the surface of the soil samples during incubation.

Our results showed that water potential within the range applied in this study is not decisive for aerobic microbial activity. However, the control of water potential during incubation is important to allow for extrapolation of the results to field conditions. The soils exhibited their maximum aerobic function at different matric potentials, which is important to modelling of plant nutrition and of gaseous and water mediated losses of nutrients from the soil. The empirical WFPS term is not able to normalize soil type differences in the water regimes of relevance to soil microbial activity.

No simple soil type independent correlation between CO2 evolution and soil water content existed in the range where water significantly increased microbial activity. The relative diffusivity of solutes calculated from the water content by recently developed models offered a better description of CO2 evolution, but further studies are needed to evaluate the mechanisms relating the relative solute diffusivity to aerobic microbial activity.

The relative gas diffusivity was a better predictor of net nitrification than was the soil air content. The results indicated that a threshold existed for the relative gas diffusivity at optimum aerobic microbial activity. This was approximately 0.025, 0.005, and 0.005 for the L1, L3, and L5 soils, respectively.

Calculations based on a conceptual model that balances the effects of solute and of gas diffusivity on aerobic microbial activity supported the relative trend in the observed optima of water contents across soil types. The modelling further confirmed a higher dominance of gas diffusivity in the sandy L1 soil compared with the more clayey L3 and L5 soils. We advocate the combined use of the conceptual model of Skopp et al. (1990) and recent soil type dependent expressions for solute and gas diffusivity (e.g., Olesen et al., 2001; Moldrup et al., 1999) in future studies of aerobic microbial activity.


   ACKNOWLEDGMENTS
 
We thank Mr. Lars Jørgen Pedersen, owner of the Lerbjerg Estate, for access to the field site, Dr. David T. Strong for supplying ceramic discs, and Drs. Lars Elsgaard and Søren O. Petersen for valuable discussions. The technical assistance of Bodil B. Christensen, Michael Koppelgaard, Stig T. Rasmussen, and Karin Dyrberg is gratefully acknowledged. This work was financially supported by the Ministry of Food, Agriculture, and Fisheries (Project HAR98-DJF-4).

Received for publication March 26, 2002.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Vadose Zone JHome page
D. A. Robinson, H. Abdu, S. B. Jones, M. Seyfried, I. Lebron, and R. Knight
Eco-Geophysical Imaging of Watershed-Scale Soil Patterns Links with Plant Community Spatial Patterns
Vadose Zone J., October 29, 2008; 7(4): 1132 - 1138.
[Abstract] [Full Text] [PDF]

Home page
 Vadose Zone JHome page
D. A. Robinson, C. S. Campbell, J. W. Hopmans, B. K. Hornbuckle, S. B. Jones, R. Knight, F. Ogden, J. Selker, and O. Wendroth
Soil Moisture Measurement for Ecological and Hydrological Watershed-Scale Observatories: A Review
Vadose Zone J., February 25, 2008; 7(1): 358 - 389.
[Abstract] [Full Text] [PDF]

Home page
 Vadose Zone JHome page
A. J.M. Smucker, E.-J. Park, J. Dorner, and R. Horn
Soil Micropore Development and Contributions to Soluble Carbon Transport within Macroaggregates
Vadose Zone J., May 17, 2007; 6(2): 282 - 290.
[Abstract] [Full Text] [PDF]

Home page
 Vadose Zone JHome page
K. Kawamoto, P. Moldrup, P. Schjonning, B. V. Iversen, D. E. Rolston, and T. Komatsu
Gas Transport Parameters in the Vadose Zone: Gas Diffusivity in Field and Lysimeter Soil Profiles
Vadose Zone J., November 20, 2006; 5(4): 1194 - 1204.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
W. S. D. Rocha, J. B. Regitano, and L. R. F. Alleoni
2,4-D Residues in Aggregates of Tropical Soils as a Function of Water Content
Soil Sci. Soc. Am. J., October 27, 2006; 70(6): 2008 - 2016.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
P. Rochette, J. Dionne, Y. Castonguay, and Y. Desjardins
Atmospheric Composition under Impermeable Winter Golf Green Protections
Crop Sci., June 20, 2006; 46(4): 1644 - 1655.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
P. Rochette, D. A. Angers, M. H. Chantigny, B. Gagnon, and N. Bertrand
In situ Mineralization of Dairy Cattle Manures as Determined using Soil-Surface Carbon Dioxide Fluxes
Soil Sci. Soc. Am. J., March 29, 2006; 70(3): 744 - 752.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
S. Agehara and D. D. Warncke
Soil Moisture and Temperature Effects on Nitrogen Release from Organic Nitrogen Sources
Soil Sci. Soc. Am. J., September 29, 2005; 69(6): 1844 - 1855.
[Abstract] [Full Text] [PDF]

Home page
 Vadose Zone JHome page
C. Kjaergaard, T. G. Poulsen, P. Moldrup, and L. W. de Jonge
Colloid Mobilization and Transport in Undisturbed Soil Columns. I. Pore Structure Characterization and Tritium Transport
Vadose Zone J., May 1, 2004; 3(2): 413 - 423.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
P. Moldrup, T. Olesen, S. Yoshikawa, T. Komatsu, and D. E. Rolston
Three-Porosity Model for Predicting the Gas Diffusion Coefficient in Undisturbed Soil
Soil Sci. Soc. Am. J., May 1, 2004; 68(3): 750 - 759.
[Abstract] [Full Text] [PDF]

This Article