HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (20) Citing Articles via Google Scholar Google Scholar Articles by Grundmann, G. L. Articles by Kizungu, R. Search for Related Content PubMed Articles by Grundmann, G. L. Articles by Kizungu, R. Agricola Articles by Grundmann, G. L. Articles by Kizungu, R.

DIVISION S-3 - SOIL BIOLOGY & BIOCHEMISTRY

Spatial Modeling of Nitrifier Microhabitats in Soil

^a Laboratoire d'Ecologie microbienne, U.M.R. C.N.R.S. 5557. UCB Lyon I, 43 Bd du 11 Novembre 1918. 69622 Villeurbanne, Cedex, France
^b Centre de Pédologie Biologique, UPR 6831 CNRS-Université Henri-Poincaré, Nancy I, 17 rue Notre Dame des Pauvres BP5, 54 501 Vandoeuvre-Les-Nancy
^c Laboratoire de Biométrie et Biologie Evolutive, U.M.R. C.N.R.S. 5558, UCB Lyon I, 43 Bd du 11 Novembre 1918, 69622 Villeurbanne, Cedex, France

* Corresponding author (grundman{at}biomserv.univ-lyon1.fr)

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Soil bacteria function in the three-dimensional space in heterogeneous soil complex and their activities depend in part on encountering substrates at the microbial scale. The bacterial density per gram of soil, which is generally measured, does not indicate if bacteria are all in the same location or spread throughout the soil complex. We characterized spatial distribution for how dispersed or aggregated nitrifiers (NH⁺₄ and NO^-₂ oxidizers) were at a submillimeter scale. The spatial approach was based on the relationship, obtained experimentally, between the percentage of microsamples (50–500 µm diam.) harboring nitrifiers and the volume of the microsamples. The smallest sample size (50-µm diam.) was considered as an approximation of microhabitat. The simulated spatial pattern of NO^-₂ oxidizer microhabitats in soil were compared with experimental data. The simulated pattern of NO^-₂ oxidizer distribution suggested that microhabitats averaged seven NO^-₂ oxidizers and occurred in preferentially colonized patches that had about a 250-µm diam. These were randomly distributed and occupied 5.5% of the soil volume. They were functionally connected through microporosity and hence diffusion processes probably controlled the spatial distribution of nirifiers. The nitrifier spatial pattern enabled efficient nitrification because NH⁺₄ and NO^-₂ oxidizers were near one another. The results showed the potential of our method to study spatial distribution of bacteria at the microhabitat scale.

Abbreviations: VU, volumetric unit • MDT, mean detection time

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
BACTERIA are responsible for major biogeochemical transformations of organic and mineral constituents in soils (Atlas and Bartha, 1981; Paul and Clark, 1989). Soil bacteria live in a complex three-dimensional habitat of a porous heterogeneous medium (Stotzky and Burns, 1982; Tisdall and Oades, 1982; Crawford and Young, 1998; Young and Ritz, 1998). The geometric complexity of soil affects the probability of bacteria encountering appropriate substrates or other bacteria. The quantitative assessment of bacteria in soil is mostly confined to population density or biomass measurements (Atlas and Bartha, 1981; Schmidt, 1982; Powlson, 1994), but rarely is the spatial organization of the cells taken into account (Hattori, 1973). For example, do bacteria in a gram of soil coalesce in a few spots or are they distributed evenly across the soil complex? Dispersion is not available for microorganisms but is routinely measured for macroscopic organisms because it determines the frequency of encountering food and other organisms. Characterizing the spatial distribution of microhabitats is important if there is to be progress in microbial ecology in soils. Also, a better understanding of the spatial arrangement of bacterial habitats should lead to the development of more appropriate bioremediation techniques (increasing probability of bacteria encountering substrates) and the optimization of soil functions (Holden and Firestone, 1997).

Bacterial activities have been reported to be unevenly distributed in soil, leading to the concept of hot spots that are linked to local, transient available C for microbial growth and activity (Parkin, 1987; Robertson et al., 1988; Beare et al., 1995). Most microbiological research is carried out on macro scales grams of soil, but bacteria cells exist and interact at the micro scale. Information of the spatial distribution of bacteria in soil is very limited, with microhabitats being poorly defined (Harris, 1994). Hattori (1973) reported results of several early studies on spatial patterns of bacteria in soil and Harris (1994) mentioned that they were mostly based on microscopic observations. The lack of quantitative data on the spatial patterns of bacteria at the microhabitat scale (Hattori, 1973, for total microflora) is because of limitations for sampling and sample processing methods.

The two main locations for active bacteria are believed to be soil pores (Hattori and Hattori, 1976; Hattori, 1988; Pievetz and Steenhuis, 1995), (within the surrounding water film), in regions of preferential flow (Harris, 1994; Kinsall et al., 1995), or alternatively entrapped within the soil matrix (Foster, 1988; Paul and Clark, 1989). These points have not yet been clearly established, nor are the mechanisms of microhabitat colonization understood.

The objective of this work was to characterize the spatial distribution of bacteria at the microhabitat scale. It was not to find the exact three-dimensional coordinates but rather to determine how dispersed or clustered bacteria were at the microscale. We used nitrifiers as a bacterial model on an undisturbed soil sample. Other research disciplines have shown that the probability of a process or organism being detected at several scales is linked to spatial patterns (Madden and Hughes, 1999). We used a microsampling procedure to measure the relationship between the percentage of the studied bacterial type present and the sample volume. A simulation procedure, projecting the soil to a three-dimensional grid was then applied to estimate the type of distribution.

Nitrifiers oxidize NH⁺₄ and NO^-₂ stepwise to produce NO^-₃ in soil. Ammonium oxidizers are obligate lithotrophs, while NO^-₂ oxidizers are mostly lithotrophic with some having heterotrophic capabilities (Schmidt, 1982; Prosser, 1989) and there is little redundancy in soil (Marilley and Aragno, 1999). The tendency of nitrifier cells to cluster in soil is based on speculation (Schmidt, 1982; Berg and Rosswall, 1986; Keen and Prosser, 1988). They are attractive to study because their substrates or products are easily measured; NO^-₂ by Griess-Ilosvay reagent (Keeney and Nelson, 1982) and H⁺ by measuring pH changes.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Preparation of Soil Volumetric Units
The agricultural soil studied was under maize (Zea Mays L.) cultivation for 10 yr. It is a sandy loam, developed from a recent glacial drift (Riss) (Typic Hapludalf) with the following characteristics: bulk density, 1.3 Mg m^-3 (Angulo et al., 1997); clay, 17.0%; silt, 39.2%; sand, 40.4%; organic C, 1.4%; pH (H₂O), 6.4 (Grundmann et al., 1995); and weakly structured (Ranjard et al., 1997). A 30-cm³ intact soil ped was taken at the 5-cm depth in June from an area without vegetation. In the laboratory, a 3-cm³ fragment was carefully taken from inside the fragment with a sterile scalpel and kept in a humid container to prevent desiccation. The soil water content at sampling was 16 g kg^-1. Over 2 d, the small soil fragment was gradually subdivided with a sterile scalpel under a binocular microscope into eight aggregates of about 0.5 cm³, to representatively sample the fragment. Each aggregate was then carefully dissected on a calibrated grid under the binocular microscope with a sterile scalpel. The calibrated grid was used to select volume units (VU) fitting into 50-, 100-, 250-, or 500-µm squares. They were considered to be cubes of 1.25 x 10^-4, 1 x 10^-3, 1.6 x 10^-2, or 0.125 mm³ volumes, which are referred to as the sizes 50, 100, 250, and 500, respectively. A series of 96 replicates of each VU size (4 VU sizes x 96 replicates) were sampled randomly from the aggregates. Although the VU dried rapidly once dissected, they were immediately placed in a culture medium to allow for bacteria survival. It was assumed that the bacteria were trapped in the dissected VU and allowed for unbiased tests of the presence or absence of the nitrifiers.

Testing for Ammonium and Nitrite Oxidizers in Volumetric Units
Each VU was taken up with a sterile plugged glass capillary that had been dipped in sterile, biologically inert silicon oil (SV40), and transferred to a 2-mL defined culture medium by swirling the tip of the capillary in the medium. Ninety-six replicates of each VU size were cultured in NH⁺₄ oxidizer culture medium (Schmidt and Belser, 1994) (1 mM NH⁺₄ final concentration) and 96 other replicates of each VU size in NO^-₂ oxidizer culture medium (Schmidt and Belser, 1994) (1 mM NO^-₂ final concentration), for 12 wk at 28°C in the dark, in the wells of micro-culture plates containing 24 wells (Schmidt and Belser, 1994). Nitrite (5 mM final concentration) was added to each culture that tested positive for NH⁺₄ oxidizers and the culture was further tested for NO^-₂ oxidizers, to assess for the presence of both NH⁺₄ and NO^-₂ oxidizers in each VU. This first series of 8 x 96 VU replicates (4 sizes for NO^-₂ and NH⁺₄ oxidizers) is referred to as Exp. A. A second set of replicates (24, 48, or 72 VU depending on VU size) was taken from the same soil clod (kept at 5°C) 2 wk later (Exp. B) (Table 1). The presence of NO^-₂ oxidizers was scored positive if there was no NO^-₂ left in the culture medium. This was determined each week by the Griess–Ilosvay spot test (Keeney and Nelson, 1982) on one drop of medium. The presence of NH⁺₄ oxidizers had a was positive score if nitrite was present or there was a drop in the pH, as tested colorimetrically on a drop of medium. All tests were carried out weekly. Ninety-six size 50 VU and 96 size 100 VU were also cultured in a nonselective medium (nutrient broth, Difco, Detroit, MI).

The approach consisted of simulating different types of distributions of positive size 50 VU taken as an elementary unit and then by simulating samplings of the distributions with the 4 experimental sizes of VU. There are many different ways to distribute positive size 50 VU in a three-dimensional grid. We have only considered homogeneous spatial patterns (i.e., one single type of spatial distribution in the whole soil volume, without any gradient). Among the three main types of point spatial distribution (Upton and Fingleton, 1985), we considered random (three-dimensional coordinates of positive size 50 VU being chosen randomly) and aggregative distributions (positive VU organized in clusters) as a regular distribution (positive VU separated by constant distances) was a priori, not realistic for describing the complex spatial pattern of NO^-₂ oxidizer microhabitats in the three-dimensional soil space. Among the numerous possible aggregative distributions, the most simple approach was to simulate aggregates around randomly distributed centers. A simulated pattern was assumed to be compatible with the experimental results when the confidence interval of the simulated results overlapped the confidence interval calculated for the experimental proportions of positive VU of each size (see below).

A flow chart of the spatial simulation is given in Fig. 1 . The soil was considered as a three-dimensional matrix (200 x 200 x 200). Each unit in the table simulated a size 50 elementary VU. The whole table thus corresponded to a 10000-µm sided cube of soil. If eijk (i = 1, 2, ..., 200, j = 1, 2, ..., 200, k = 1, 2, ..., 200) was a unit in the three-dimensional table then eijk = 1 when there were bacteria in this simulated soil volume unit or if eijk = 0 when bacteria were absent. Since only the presence or absence of nitrifiers was determined, eijk = 1 gives no indication of population size. The proportion of eijk = 1 in the grid corresponds to the density of positive elementary VU in the grid, noted d. The range of densities, d, tested was [0.001, 0.15] with 0.001 increments. The simulation programs were written in C language and calculations were carried out on a Sun station (Sun Microsystems, Palo Alto, CA) using a mixed congruential pseudorandom number generator: Xn + 1 = bXn + g (modulo m) with b = 69069, g = 1, and m = 2³².

View larger version (57K): [in this window] [in a new window]   Fig. 1. Flow chart of simulation procedure for the spatial pattern of nitrifier microhabitats. The two examples represent a section of the three-dimensional simulation for one value of d. The letter d denotes density of units taking the value one (bold character for random distribution, noted as 1) or of size 50 elementary unit harboring at least one bacteria. The letter a denotes size of the aggregate side (number of size 50 elementary units) when Q = 1. The initial positive VU is the center of the aggregate (noted 1) (a = 3 or 5). Q: Probability that each adjacent unit in an aggregate belongs to same aggregate (Q = 0.3 or 0.7).

Sampling Simulation of the Grid
To mimic the actual experiment, each simulated grid, characterized by a density of positive elementary units (d) and a type of spatial distribution, was sampled with 100 samples of each sample volume. The sample volumes were in the form of cubes of side c, with c = 1, 2, 5, or 10 units of the table so as to simulate sizes 50, 100, 250, and 500 of the experimental sampling.

Comparison of Simulated and Experimental Results
The soil was considered to be a three-dimensional structure containing positive or negative VUs of size c with the unknown probabilities, p_c and 1-p_c, respectively; allowing the confidence interval on the experimental proportion of positive detections to be calculated as follows below. Only a fraction K of n VU of one size gave subsequent growth. As the binomial distribution B{n, p_c} is the distribution of the number of successes that occur in n independent trials, where the probability of success in each trial is p_c (the proportion of positives) and K/n is an estimate of p_c. The ratio K/n is issued from an unknown binomial distribution belonging to a set of binomial distributions B{n,p_c} which contain K/n in the interval formed by their 5 and 95% quantiles (P > 0.05). The highest and lowest limits of the p_c values of these distributions were taken as an estimator of the confidence limits of the observed K/n. For instance, if n = 96 and K = 31, the observed probability (0.32) may in fact be derived from binomial distributions, where p_c ranged from 0.232 to 0.426. Four confidence intervals on the four values of p_c (one for each VU size) were obtained. For the simulated results, a 95% confidence interval was built on the mean of the 300 repetitions, as the mean is an unbiased estimator of a binomial distribution parameter. It seems reasonable that if the simulated spatial distribution is not too different from the distribution in soil, then the four confidence intervals on p_c and their corresponding intervals from the simulation should overlap, thus establishing agreement between experimental and simulated results. This does not constitute a statistical test, and agreement simply means that the initial distribution in the soil could be of the same type as the simulated one. Thus, this procedure enabled elimination of the most unlikely distributions. Simulated data was compared with experimental data on NO^-₂ oxidizers in Exp. A shown in Table 1.

Descriptions of Bacterial Spatial Patterns
The results of the simulation (i.e., the relevant aggregate size a (preferentially colonized patches) and the corresponding d) were then used to estimate the average distance, L, between two aggregate gravity centers. This was done by transforming the formula used to calculate the average distance, L, between the centers of particles in a volume, V, containing N particles:

[1]

If V is divided between k unit volumes, u, corresponding to the relevant size of preferentially colonized patches, and if N represents the number of preferentially colonized patches (of c³ units harboring at least one bacteria), then, for regular aggregates, L can be written as:

[2]

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Simulation of Nitrifier Microsite Spatial Patterns
The probability, p_c, of obtaining positive VU in each VU size was estimated by the observed proportion of VU harboring NO^-₂ oxidizers or NH⁺₄ oxidizers. This probability dropped from about one at size 500 down to near zero at size 50 (Table 1). All samples corresponded to the random exploration of about 56 x 10^-4 cm³ of the initial 3 cm³ soil volume (i.e., 0.2%). The sampling size threshold for which some VU did not harbor bacteria growing on nutrient broth was size 100 (98% positive size 100, 73% size 50). The existence of VU that yielded no bacterial growth in the broad spectrum medium indicates that the sampling procedure did not cause general contamination of VU.

The comparison between simulated and experimental results of Exp. A on NO^-₂ oxidizers showed that the spatial distribution of positive size 50 elementary VU was aggregated. The simulated probabilities of the four VU sizes harboring nitrifiers tested were not in agreement with the experimental data when the distribution was not aggregated. They did agree with the experimental data (confidence interval overlapped) in the case of irregular aggregates of side a = 5 (<250 µm) with an erosion probability Q = 0.2, coupled with the range of density values from 5.2 to 5.9% (Fig. 2) . Using Eq. [2] the average distance between the centers of colonized patches was estimated as 375 µm.

View larger version (28K): [in this window] [in a new window]   Fig. 2. Graph of agreement zone between experimental and simulated results from the aggregated distribution a = 5, Q = 0.2. The confidence interval around simulated probabilities for a size c VU that is positive as a function of d (d [0.1%, 12%]) is indicated by dotted lines. These were obtained by virtual sampling (c = 1, 2, 5, 10) of 300 grids issued from an aggregated distribution a = 5, Q = 0.2. The confidence interval around the experimental values is indicated by horizontal lines and brackets. The agreement zone (d [5.2, 5.9]) between the simulated and experimental results is indicated by the vertical lines.

View larger version (72K): [in this window] [in a new window]   Fig. 3. An example of a simulated spatial pattern for NO-2 oxidizer colonized patches obtained from the random distribution of irregular aggregates (d = 5.5%, a = 5 and probability of positive cubes in an aggregate Q = 0.2). Each cube (50-µm side) harbored at least one bacterium. The whole simulation is for a 1-mm side cube of soil.

Relative Distribution of Nitrite and Ammonium Oxidizers
Table 1 shows that VU positive for NO^-₂ oxidizers in Exp. A were more frequent than for NH⁺₄ oxidizers when considering VUs larger than size 50 (this size would require a larger number of samples because of the low number of positive readings). This was not the case for size 250 when tested for the simultaneous presence of both bacterial types because 48% of the readings were positive for NO^-₂ oxidizers which was obviously underestimating for unknown reasons. In another independent experiment (Grundmann and Debouzie, 2000), VU positive for NO^-₂ oxidizers were also more frequent than for NH⁺₄ oxidizers. The fact that in four independent cases, NO^-₂ oxidizers tended to be more frequent than NH⁺₄ oxidizers would indicate that the spatial pattern of the NO^-₂ oxidizers was more diffuse than the NH⁺₄ oxidizers. The percentage of nitrifiers present in the various VU sizes in Exp. B, which is a repeat of Exp. A, behaved similarly, but were higher than in Exp. A, indicating that spatial modification probably occurred during the 15-d storage at 4°C, which may be due to heterogeneity or changes in population density.

The simultaneous presence of NO^-₂ and NH⁺₄ oxidizers in VU was studied on size 100 (Table 1) to focus on a fine enough spatial scale of bacterial colonies relevant to microhabitats. A ² test on the 96 size 100 VU in Exp. A (Table 1) showed that the NO^-₂ and NH⁺₄oxidizers were not independent (P < 0.05). This indicated that at this submillimeter sampling scale, NO^-₂ oxidizers tended to be spatially associated with NH⁺₄ oxidizers. This could not be shown on size 250 because of the anomalously low level of NO^-₂ oxidizers. In another experiment by Grundmann and Debouzie (2000), a spatial association was shown for size 250. Of course, the larger the sample size above a threshold, the higher the probability to have both bacterial types in a sample. This was clearly shown for size 500. Association of both bacterial types for small sample sizes thus defines spatial functional units where nitrite is both produced and used at very short distance.

Kinetics of Nitrite Oxidizers
The number of positive VU, K, increased with time for each VU size and reached a maximum. The K_max values and curve shapes seemed to reveal a different activity behavior among the different sizes (Fig. 4a) . The K/K_max ratio was plotted against time (Fig. 4b). The calculated error did not take the sampling error of K_max into account. The K/K_max vs time curves were sigmoidal (S-shaped), which was expected when the density probability has a normal distribution. The number of positive detections appearing at a given time for each VU size, increased to a maximum. This point in time corresponded to the mean of the number of days of incubation needed to obtain a positive VU or mean detection time (MDT). After that, it decreased steadily. The MDT was 33.9 d for size 500, 46.3 d for size 250, and 53 d for size 100. MDT decreased as the soil volume increased. The simplest explanation was that the variation in MDT was essentially because of the different number of bacteria initially present in the soil sample and not to the heterogeneity of their physiological status. The mean lag phase and the mean growth rate were thus expected to be the same for all bacteria, whatever sample size they came from.

View larger version (28K): [in this window] [in a new window]   Fig. 4. (a) Kinetics of positive NO-2 oxidizers (proportion) cultures in VU sizes 100, 250, or 500 and (b) normalized kinetics for the presence of positive VUs at a given time versus the number of maximum positive VU obtained at the end of the incubation, as a function of time.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

The size 50 elementary VU of Fig. 3, was greater than the cell occupancy volume, so that the observed spatial pattern does not represent the bacterial pattern sensu stricto. Positive elementary VU may include: (i) nitrifier microcolonies or isolated cells; (ii) their sensu stricto microhabitats (surrounding environment); and (iii) the embedded soil microvolumes around these soil microhabitats. Thus, spots of NO^-₂ oxidizers were quite remote from each other as the average distance between preferentially colonized locations was relatively large at 375 µm.

This seems paradoxical because nitrification is very rapid in this soil (Grundmann et al., 1995). This requires spatial and temporal synchrony of substrate and nitrifying bacteria. The high nitrification rates can be explained by NH⁺₄ and NO^-₂ oxidizers not being independently located and that NO^-₂ oxidizers were more spatially diffuse than NH⁺₄ oxidizers. This spatial pattern could reflect an efficient NO^-₂ interception strategy by NO^-₂oxidizers to capture the soluble and mobile NO^-₂ ion. In situ, hybridization experiments on sludge flocs have revealed that Nitrobacter and Nitrosomonas species occurred in clusters and frequently were in contact with each other (Mobarry et al. 1996). Our results indicate that nitrifiers would behave similarly in soil.

Microporosity was probably the only porosity in VU (bulk density of aggregates was 1.8 Mg m^-3 and 1.3 Mg m^-3 for the undisturbed soil). The main pore volume for 500 VU, as determined by mercury porosimetry, corresponded to pore radii of 0.05 to 3.4 µm (data not shown). Fracture lines probably formed along large pores when soil was dissected, and we do not know the position of VU in relation to macro and microporosity. Cells situated on the surface of VU could have been associated with larger pores.

The precision of the proposed spatial descriptors (the diameter of preferentially colonized patches and percentage occupancy) and the validity of inferences based on these descriptors depends on the ability of the sampling technique to represent reality and on the simulation procedure. This is why we determined the confidence intervals for both experimental and simulated values. The larger the number of sample sizes and simulated distributions studied, the greater the reliability of inferences based on the descriptors. In our model, heterogeneity of the spatial distribution of bacteria in soil was considered to be a random spatial process, governed by probability. However, heterogeneity that may have once been defined as random variation may be found to contain a systematic component as soil is studied in greater detail (Trangmar et al., 1985). The mechanism that controls NH⁺₄ distribution in soil has not been clearly established.

Underhill and Prosser (1987) mentioned that the substrate (NH⁺₄, NO^-₂) adsorption site was the major factor in the attachment and colonization of nitrifiers. Papendick and Campbell (1981) indicated that distribution of NH⁺₄ was affected by its diffusion in the soil water film. This agrees with the observed tendency for both NH⁺₄ and NO^-₂ oxidizers to be present in VU only short distances apart. If spatial relationships were based on the main fluxes of substrates, the need to be close together would be weaker.

The fact that all the VU sizes had the same rate of positive appearance on normalized curves (Fig. 4) suggests that the activity of VU depends on the number of cells only, and that cell clusters of various sizes are evenly distributed in colonized patches of soil. This interpretation has different consequences from one that assumes that cells are in different physiological states. Therefore, even remote cells, far from the main fluxes could be active when the microenvironmental conditions are favorable. This agrees with Jensen et al. (1993) who reported an immediate activation of nitrifying bacteria after elimination of adverse conditions. This result provides further support for the approach adopted by Molina (1985) and Darrah et al. (1987), who considered that the rate of nitrate production varied with the relative size of cell clusters and surrounding soil spheres. They emphasized that self-inhibition by acid production was a controlling factor, but our work shows the importance of substrate diffusion processes.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
The nitrifier, microhabitat spatial pattern was made up of randomly distributed preferentially colonized patches within the soil matrix. Nitrite and NH⁺₄ oxidizers were not spatially independent, and NO^-₂ oxidizers were distributed more widely to efficiently intercept NO^-₂ a water soluble and mobile ion. This indicates a nonrandom distribution of cells within the preferentially colonized patches of soil. Because of the distance between NO^-₂ oxidizer colonized patches, soil function must be based on a high intrinsic physicochemical connectivity in the matrix, particularly if micropores are the main microhabitat location.

It is crucial to determine the relationship that may exist between the spatial patterns of bacterial microhabitats or microcolonies and the size, heterogeneity, and spatial pattern of the soil pore networks, particularly those which are bacterial microhabitats. An understanding of bacterial microhabitat distribution would mean that a large amount of existing information on soil physical and chemical properties could be used to better understand microbial ecology.

	ACKNOWLEDGMENTS

 
We thank M. Kiensele for help with the modeling and J.P. Gaudet, P.J. Harris, L. Jocteur-Monrozier and Y. Moënne-Loccoz, for their comments on various parts of this manuscript.

Received for publication March 18, 2001.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Angulo-Jaramillo, R., F. Moreno, B.E. Clothier, J.L. Thony, G. Vachaud, E. Fernandez-Boy, and J.A. Cayuela. 1997. Seasonal variation of hydraulic properties of soils measured using a tension disk infiltrometer. Soil Sci. Soc. Am. J. 61:27–32.
• Atlas, R.M., and R. Bartha. 1981. Microbial Ecology. Fundamentals and applications. Addison-Wesley, Reading, MA.
• Beare, M.H., D.C. Coleman, D.A. Crossley, P.F. Hendrix, and E.P. Odum. 1995. A hierarchical approach to evaluating the significance of soil biochemical cycling. Plant Soil 170:5–22.
• Berg, P., and T. Rosswall. 1986. Seasonal variations in abundance and activity of nitrifiers in four arable cropping systems. Microb. Ecol. 13:75–87.
• Crawford, J.W., and I.M. Young. 1998. The interactions between soil structure and microbial dynamics. p. 233–260. In P. Baveye et al. (ed.) Fractals and chaos in soil science. Advances in Soil Science. CRC Press LLC, Boca Raton, FL.
• Darrah, P.R., R.E. White, and P.H. Nye. 1987. A theoretical consideration of the implication of cell clustering for the prediction of nitrification in soil. Plant Soil 99:387–400.
• Foster, R.C. 1988. Micro-environments of soil micro-organisms. Biol. Fertil. Soils 6:189–203.
• Grundmann, L.G., P. Renault, L. Rosso, and R. Bardin. 1995. Differential effects of soil water content and temperature on nitrification and aeration. Soil Sci. Soc. Am. J. 59:1342–1349.[Abstract/Free Full Text]
• Grundmann, G.L., and D. Debouzie. 2000. Geostatistical analysis of the distribution of NH⁺₄ and NO^-₂ oxidizing bacteria and serotypes at the millimetric scale along a soil transect. FEMS Microbiol. Ecol. 1170:1–6.
• Harris, P.J. 1994. Consequences of spatial distribution of microbial communities in soil. p. 239–247. In K. Ritz (ed.) Beyond the biomass. John Wiley and Sons, Chichester, UK.
• Hattori, T. 1973. Microbial life in the soil. Marcel Dekker, New York.
• Hattori, T., and R. Hattori. 1976. The physical environment in soil microbiology: An attempt to extend principles of microbiology to soil microorganisms. Crit. Rev. Microbiol. 4:423–460.
• Hattori, T. 1988. Soil aggregates as microhabitats of microorganisms. Rep. Inst. Agric. Res., Tohoku Univ. 37:23–36.
• Holden, P.A., and M.K. Firestone. 1997. Soil microorganisms in soil clean-up: How can we improve our understanding? J. Environ. Qual. 26:32–40.[Abstract/Free Full Text]
• Jensen, K., N.P. Revsbeck, and L.P. Nielsen. 1993. Microscale distribution of nitrification activity in sediment determined with a shielded microsensor for nitrate. Appl. Environ. Microbiol. 59:3287–3296.[Abstract/Free Full Text]
• Keen, G.A., and J.I. Prosser. 1988. The surface growth and activity of Nitrobacter. Microbiol. Ecol. 15:21–39.
• Keeney, D.R., and D.W. Nelson. 1982. Nitrogen—Inorganic forms. p. 643–698. In A.L. Page et al. (ed.) Methods of soil analysis. Part 2. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
• Kinsall, B.L., G.V. Wilson, A.V. Palumbo, and T. Phelps. 1995. Correlation of heterogeneity in physical and chemical properties of the vadose zone with microbe populations and migration. p. 243. In 1995 Agronomy abstracts. ASA, Madison, WI.
• Madden, L.V., and G. Hughes. 1999. An effective sample size for predicting plant disease incidence in a spatial hierarchy. Phytopathology 89:770–781.
• Marilley, L., and M. Aragno. 1999. Phylogenetic diversity of bacterial communities differing in degree of proximity of Lolium perenne and Trifolium repens roots. Appl. Soil Ecol. 13:127–136.
• Mobarry, B.K., M. Wagner, V. Urbain, B.E. Rittman, and D.A. Stahl. 1996. Phylogenetic probes for analyzing abundance and spatial organization of nitrifying bacteria. Appl. Environ. Microbiol. 62: 2156–2162.[Abstract]
• Molina, J.A.E. 1985. Components of rates of ammonium oxidation in soil. Soil Sci. Soc. Am. J. 49:603–609.[Abstract/Free Full Text]
• Nishio, M., and C. Furusaka. 1970. The distribution of nitrifying bacteria in soil aggregates. Soil Sci. Plant Nutr. 16:24–29.
• Papendick, R.I., and G.S. Campbell. 1981. Theory and measurement of water potential. p. 1–22. In J.F. Parr et al. (ed.) Water potential relations in soil microbiology. SSSA Spec. Publ. 9. SSSA, Madison, WI
• Parkin, T.B. 1987. Soil microsites as a source of denitrification variability. Soil Sci. Soc. Am. J. 51:1194–1199.[Abstract/Free Full Text]
• Paul, E.A., and F.E. Clark. 1989. Soil microbiology and biochemistry. Academic Press, San Diego, CA.
• Pivetz, B.E., and T.S. Steenhuis. 1995. Soil matrix and macropore biodegradation of 2,4-D. J. Environ. Qual. 24:564–570.[Abstract/Free Full Text]
• Powlson, D.S. 1994. The soil microbial biomass: before, beyond and back. p. 3–19. In K. Ritz et al. (ed.) Beyond the biomass. John Wiley and Sons, Chichester, UK.
• Prosser, J.I. 1989. Autotrophic nitrification in bacteria. Adv. Microb. Physiol. 30:125–181.[Medline]
• Ranjard, L., A. Richaume, L. Jocteur-Monrozier, and S. Nazaret. 1997. Response of soil bacteria to Hg(II) in relation to soil characteristics and cell location. FEMS Microbiol. Ecol. 24:321–331.
• Robertson, G.P., M.A. Huston, and F.C. Evans. 1988. Spatial variability in a successional plant community: patterns of nitrogen availability. Ecology 69:1517–1524.[ISI]
• Schmidt, E.L. 1982. Nitrification in soil. p. 253–288. In F.J. Stevenson (ed.) Nitrogen in agricultural soils. Agron. Monogr. 22. ASA, CSSA, and SSSA, Madison, WI.
• Schmidt, E.L., and L.W.Belser. 1994. Autotrophic nitrifying bacteria. p. 159–177. In R.W. Weaver et al. (ed.) Methods of soil analysis. Part 2. SSSA Book Ser. 5. SSSA, Madison WI.
• Stotzky, G., and R.G. Burns. 1982. The soil environment: Clay-humus-microbe interactions. p. 105–133. In R.G. Burns and J.H. Sloter (ed.) Experimental microbial ecology. Blackwell Scientific Publications, Oxford.
• Tisdall, J.M., and J.M. Oades. 1982. Organic matter and water stable aggregates in soil. J. Soil Sci. 32:141–163.
• Trangmar, B.B., R.S. Yost, and G. Uehara. 1985. Application of geostatistics to spatial studies of soil properties. Adv. Agron. 38:45–94.
• Underhill, S.E., and J.I. Prosser. 1987. Surface attachment of nitrifying bacteria and their inhibition by potassium ethyl xanthate. Microb. Ecol. 14:129–139.
• Upton, G., and B. Fingleton. 1985. Spatial data analysis by example. Vol. 1. Point pattern and quantitative data. John Wiley and Sons, Chichester, UK.
• Young, I.M., and K. Ritz. 1998. Can there be a contemporary ecological dimension to soil biology without a habitat? Soil Biol. Biochem. 30:1229–1232.

This article has been cited by other articles:

				J. M. Norton, M. G. Klotz, L. Y. Stein, D. J. Arp, P. J. Bottomley, P. S. G. Chain, L. J. Hauser, M. L. Land, F. W. Larimer, M. W. Shin, et al. Complete Genome Sequence of Nitrosospira multiformis, an Ammonia-Oxidizing Bacterium from the Soil Environment Appl. Envir. Microbiol., June 1, 2008; 74(11): 3559 - 3572. [Abstract] [Full Text] [PDF]

				H. Sanguin, B. Remenant, A. Dechesne, J. Thioulouse, T. M. Vogel, X. Nesme, Y. Moenne-Loccoz, and G. L. Grundmann Potential of a 16S rRNA-Based Taxonomic Microarray for Analyzing the Rhizosphere Effects of Maize on Agrobacterium spp. and Bacterial Communities. Appl. Envir. Microbiol., June 1, 2006; 72(6): 4302 - 4312. [Abstract] [Full Text] [PDF]

				L. Vieuble Gonod, J. Chadoeuf, and C. Chenu Spatial Distribution of Microbial 2,4-Dichlorophenoxy Acetic Acid Mineralization from Field to Microhabitat Scales Soil Sci. Soc. Am. J., December 2, 2005; 70(1): 64 - 71. [Abstract] [Full Text] [PDF]

				A. Dechesne, C. Pallud, F. Bertolla, and G. L. Grundmann Impact of the Microscale Distribution of a Pseudomonas Strain Introduced into Soil on Potential Contacts with Indigenous Bacteria Appl. Envir. Microbiol., December 1, 2005; 71(12): 8123 - 8131. [Abstract] [Full Text] [PDF]

				C. Pallud, A. Dechesne, J. P. Gaudet, D. Debouzie, and G. L. Grundmann Modification of Spatial Distribution of 2,4-Dichlorophenoxyacetic Acid Degrader Microhabitats during Growth in Soil Columns Appl. Envir. Microbiol., May 1, 2004; 70(5): 2709 - 2716. [Abstract] [Full Text] [PDF]

				J. Vogel, P. Normand, J. Thioulouse, X. Nesme, and G. L. Grundmann Relationship between Spatial and Genetic Distance in Agrobacterium spp. in 1 Cubic Centimeter of Soil Appl. Envir. Microbiol., March 1, 2003; 69(3): 1482 - 1487. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (20) Citing Articles via Google Scholar Google Scholar Articles by Grundmann, G. L. Articles by Kizungu, R. Search for Related Content PubMed Articles by Grundmann, G. L. Articles by Kizungu, R. Agricola Articles by Grundmann, G. L. Articles by Kizungu, R.