SEARCH

SEARCH BY CITATION

Keywords:

  • flagellates;
  • flagellate growth;
  • Cercomonas sp.;
  • actinobacteria;
  • food quality of bacteria;
  • prey selectivity

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion and perspectives
  8. Acknowledgements
  9. References

Flagellates are very important predators on bacteria in soil. Because of their high growth rates, flagellate populations respond rapidly to changes in bacterial numbers. Previous results indicate that actinobacteria are generally less suitable than proteobacteria as food for flagellates. In this study, we investigated the growth of the flagellate Cercomonas sp. (ATCC 50334) on each of the two bacteria Sphingopyxis witflariensis (Alphaproteobacteria) and Rhodococcus fascians (actinobacteria) separately and in combination. The growth rate of the flagellate was lower and the lag phase was longer when fed with R. fascians than when fed with S. witflariensis. This supports our initial hypothesis that the actinobacterium is less suitable as food than the alphaproteobacterium. However, after longer periods of growth the peak abundance of flagellates was higher on R. fascians, indicating that the food quality of bacterial prey depends on the time perspective of the flagellate–bacterial interaction. There was no evidence that the flagellates selected against the actinobacterium when feeding in mixed cultures of the two bacteria. Experiments where flagellates were fed with washed bacterial cells or with bacteria growing with different substrate concentrations suggested that the low food quality of R. fascians is related both to the intrinsic cell properties and to the extracellular metabolites.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion and perspectives
  8. Acknowledgements
  9. References

Protozoa are important grazers of bacteria in soil (Ekelund & Rønn, 1994). The grazing activity of protozoa stimulates bacterially mediated processes such as mineralization (Hunt et al., 1987; deRuiter et al., 1993; Ekelund & Rønn, 1994) and nitrification (Griffiths, 1989; Verhagen et al., 1993) and can change the composition of bacterial communities in soil (Griffiths et al., 1999; Rønn et al., 2002). The mechanisms that lead to a change in bacterial communities as a result of protozoan predation are not clear but studies from aquatic systems have shown that protozoa may feed selectively on different bacteria according to their size (Chrzanowski & Šimek, 1990; Gonzalez, 1996; Kinner et al., 1998; Posch et al., 1999). Many flagellates and ciliates preferentially consume intermediate-sized bacteria (Hahn & Höfle, 2001). Other bacterial properties such as chemical composition (Jürgens & DeMott, 1995), cell surface properties (Wootton et al., 2007), cell motility (Matz & Jürgens, 2003) and microcolony formation (Matz et al., 2002) have also been found to affect the likelihood that bacterial cells are consumed by protozoa. Using FISH, Jezbera et al. (2005) observed a negative selection of actinobacteria by protozoa in a freshwater system, although this may partly have been a consequence of size-selective grazing (Jezbera et al., 2005; Šimek et al., 2005).

The food quality of different bacteria for protozoa varies considerably. Pigmentation and production of various intra- and extra-cellular metabolites can affect protozoan growth (Singh, 1945; Groscop & Brent, 1964; Weekers et al., 1993; Andersen & Winding, 2004; Matz et al., 2004; Jousset et al., 2006). Some researchers have observed a relationship between the Gram status of bacteria and their food quality for protozoa. Chang (1960) observed no growth of three different amoebae on two strains of low G/C Gram-positive bacteria (both Bacillus sp.). Weekers et al. (1993) found a more limited growth of three species of amoeba when fed with the high G/C Gram-positive bacterium Micrococcus luteus compared with two low G/C Gram-positive bacteria and two strains of nonpigmented proteobacteria. Mohapatra & Fukami (2005) showed a slower growth of the flagellate Jakoba libera when fed with Micrococcus sp. compared with when it was fed with either one of two strains of Gram-negative bacteria (Pseudomonas sp. or Vibrio sp.) or a filtrate of marine bacteria. Bjørnlund et al. (2006) observed a very limited (or no) growth of the flagellate Cercomonas sp. when supplied with a number of different strains of actinobacteria (high G/C Gram-positive), whereas they generally found a relatively greater growth on various proteobacteria (mostly Alphaproteobacteria).

These results indicate that the composition of bacterial communities affects the growth and abundance of flagellate predators. In natural systems, bacteria that are more or less nutritious and/or palatable usually occur together. Therefore, it is relevant to understand how flagellates respond to mixed communities of bacteria. However, most detailed growth experiments have been performed with pure cultures of single bacteria. The aim of the present study was to examine how flagellate growth is affected by different combinations of bacteria with high and low food quality, respectively. We investigated the growth of the heterotrophic flagellate Cercomonas sp. on two different strains of bacteria. The genus Cercomonas is very common in soil (Ekelund et al., 2001) and it is therefore relevant to use a member of this genus as a model organism (Bjørnlund et al., 2006). The strain of Cercomonas sp. is identical to the axenic strain (ATCC 50334) used by Bjørnlund et al. (2006). We selected two of the bacterial strains from the study by Bjørnlund et al. (2006) based on their ability to support flagellate growth. We chose one alphaproteobacterium (Sphingopyxis witflariensis) giving rise to substantial flagellate growth and one actinobacterium (Rhodococcus fascians) that supported virtually no growth of Cercomonas sp. in growth experiments lasting 5 days (Bjørnlund et al., 2006). Based on the previous observations with these two bacteria and the literature reports of differences in food quality between Gram-positive and Gram-negative bacteria, we expected flagellate growth to be severely limited when fed with R. fascians and we hypothesized that the lower food quality of R. fascians is due to properties related to the nature of the Gram-positive cell wall. Furthermore, we hypothesized that the flagellate would feed selectively on ‘suitable’ food bacteria. We carried out experiments to test the following hypotheses: (1) the growth rate and peak abundance of Cercomonas sp. is higher on S. witflariensis than on R. fascians; (2) the presumed difference in food quality of the two bacteria is related to a difference in the composition of the bacterial cells (not due to extracellular metabolites); and (3) under combined culture conditions, Cercomonas sp. feeds selectively on S. witflariensis and avoids R. fascians. We also investigated how the growth of the flagellate was affected by different combinations of the bacteria compared with the growth on either of the two bacteria separately. To meet these goals, we set up a number of different growth experiments where the flagellates were fed with washed bacterial cells or unwashed bacterial cultures of the two bacteria alone and in different combinations.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion and perspectives
  8. Acknowledgements
  9. References

Organisms

The strain of Cercomonas sp. used in this study was obtained as an axenic culture from the American Type Culture Collection (ATCC 50334) and is maintained at 15 °C on a mixture of heat-killed cells of the gammaproteobacterium Enterobacter aerogenes [SC40, provided by Søren Christensen (Christensen & Bonde, 1985)] and a nutrient medium (axenic Dimastigella medium – ATCC medium 1865). The flagellate cells are c. 5 μm long in the amoeboid phase (width: 2.8 μm). The two strains of bacteria used in this experiment were originally isolated from barley rhizosphere soil (Vestergård et al., 2004) and are identical to the strains BEM660 (an actinobacterium) and BEM760 (an alphaproteobacterium) used in a previous study (Bjørnlund et al., 2006). The strains were sent to the identification service of Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ) in Braunschweig, Germany, and identified as R. fascians (BEM 660) and S. witflariensis (BEM 760), respectively. The Rhodococcus strain is included in the DSMZ collection under the accession number DSM 45106. The two strains have distinct colony morphologies. The cells of both strains are rod shaped, R. fascians being larger than S. witflariensis (Table 1).

Table 1.   Data for the two bacterial strains used
StrainIdentification*Shape of cellsLength (μm) (SE)Width (μm) (SE)Volume (μm3)Colony morphologyGenBank accession no.
  • *

    Identifications performed by DSMZ (Braunschweig, Germany).

BEM660Rhodococcus fasciansRods1.6 (0.046)0.79 (0.018)0.65Orange, smoothAM110769
BEM760Sphingopyxis witflariensisRods1.2 (0.028)0.57 (0.011)0.25Yellow, smoothAM110780

Preculturing of bacteria

Before the experiments, the bacteria were transferred from agar plates to Erlenmeyer flasks containing tryptic soy broth (TSB, 3 g L−1) dissolved in AS (modified Neff's amoeba saline; Page, 1988). These precultures were incubated on a shaker (Gerhard LS500, 70 strokes min−1) in a climate chamber at 15 °C in darkness.

Experimental arrangement

Either washed bacterial cells resuspended in AS or bacterial cultures grown at different concentrations of TSB were pipetted into 25 cm2 cell culture flasks (Nunc, Intermed). The flagellates were subsequently added to the flasks. We performed repeated, direct counting on the same microcosms and growth parameters such as growth rate were calculated for the individual flasks. Growth was generally followed until the stationary phase was reached. At appropriate times, samples were taken out for estimation of bacterial numbers by plate counts. All incubations were carried out in a climate chamber at 15 °C.

Experiment 1: Growth on washed bacteria in different ratios

Five cultures of the two bacteria were prepared by adding 500 μL of the two precultures to 500-mL Erlenmeyer flasks, each containing 200 mL of 0.3 g TSB L−1 in AS. The flasks were placed on a shaker in a climate chamber at 15 °C. The bacteria were allowed to reach the stationary phase, evaluated via the OD of the cultures and centrifuged at 6660 g for 10 min in 200-mL centrifuge tubes. The supernatant was discarded and 100 mL of AS was added to each tube. After vortex mixing of the contents, the tubes were centrifuged again. This washing and centrifugation procedure was repeated three times. The bacteria were resuspended in AS and the density of cells in the suspensions was estimated by direct counting using epifluorescence microscopy. Each sample was stained with 1 mL of sterile-filtered acridine orange solution and 3 mL of sterile-filtered acetic acid solution (600 μL 99.9% acetic acid L−1) was added as a buffer. After 2 min, the suspension with stained bacteria was sucked down onto a 0.2 μm Nuclepore carbon filter (Corning Incorporated, NY) and the filter was washed with 2 mL of sterile-filtered acetic acid solution to remove any access dye. The filter was mounted with paraffin oil on a microscope slide. The number of bacteria in 25 counting grids was counted at × 1250 magnification using an Olympus BH2 microscope. Duplicate filters were counted for both bacteria. Based on these counts, we diluted the culture of S. witflariensis 1.25 × and the R. fascians culture 0.5 × (1 mL of the original culture corresponded to 0.5 mL of the final resuspension). This was done in order to obtain the same cell density in the two bacterial suspensions. A total volume of 5-mL bacterial suspension was added to cell culture flasks. The two bacterial suspensions were added in different amounts to obtain the following estimated ratios of bacterial cells in the flasks (R. fascians/S. witflariensis): 100/0, 90/10, 70/30, 50/50, 30/70, 10/90 and 0/100. Six flasks were prepared with each bacterial ratio. A small inoculum of Cercomonas sp. from the axenic culture was added to three flasks for each bacterial ratio. The same volume of AS was added to an equal number of control flasks. The flasks were incubated at 15 °C and the flagellates were counted at intervals as described below. After 11 days, we estimated the number of colony-forming units (CFU) in treatments 50/50 and 0/100. A subsample (100 μL was taken from each flask and 10-fold dilution series were prepared by diluting with AS. Appropriate dilutions (10−5, 10−6 and 10−7) were plated in duplicate on agar plates (3 g TSB L−1, 15 g Agar L−1 dissolved in AS). The plates were incubated at 15 °C and counted after 3 weeks. At Day 21, we estimated the number of CFUs in five treatments (100/0, 70/30, 50/50, 30/70 and 0/100) following the same procedure as for Day 11.

Experiment 2: Growth on washed bacteria and with fresh TSB medium

In the first experiment, we observed a delayed growth response of the flagellates on R. fascians. In order to test whether this phenomenon was due to a change in bacterial cells resulting from the prolonged starvation, we set up an experiment where fresh bacterial growth medium was added to the experimental microcosms at intervals during the experiment. Bacteria were cultured and harvested following the procedure outlined for the washed bacteria experiment. The bacteria were resuspended in the same volume of AS as the culture they were grown in. A 50/50 mixture of the two bacteria was prepared. For each bacterial suspension six cell culture flasks were filled with 5 mL of the suspension. All flasks were inoculated with c. 100 flagellate cells mL−1. On Days 5, 8 and 23, half of the flasks were supplied with 100 μL of fresh TSB medium (3 g TSB L−1 in AS), and the other half with 100 μL of a 10 × concentrated suspension of the bacterial culture.

Experiment 3: Growth on unwashed bacterial culture

This experiment was carried out three times in total with slight modifications but only results from the third of these experiments, where a mixture of the two bacteria was included as a treatment, are shown here. Subsamples (10 μL) of the R. fascians or S. witflariensis preculture were inoculated into cell culture flasks containing 5 mL of the following TSB solutions (in AS): 1.2, 0.6, 0.3, 0.2, 0.15, 0.1, 0.075, 0.06, 0.04 and 0.03 g L−1. Six flasks with each bacterium were prepared for each of the 10 TSB concentrations. The cultures were allowed to reach the stationary phase. The contents from one of the flasks with 1.2 g TSB L−1R. fascians culture were mixed and poured into one of the flasks with 1.2 g TSB L−1S. witflariensis culture. This mixture (c. 10 mL) was mixed thoroughly and poured back into the now empty culture flask. The contents were mixed thoroughly and 5 mL of the bacterial suspension was pipetted into a new cell culture flask. This procedure was followed for triplicate bottles of all TSB concentrations. An inoculum of c. 100 Cercomonas sp. cells from a 14-day-old culture was added to a total of 90 cell-culture flasks with bacterial cultures (30 with Sphingopyxis, 30 with Rhodococcus and 30 with the bacterial mixture).

Experiment 4: Large inoculum size

In Experiment 3, we observed a pronounced lag phase when flagellates were growing with R. fascians as food. The length of the lag phase was longer with higher medium concentrations. This led us to speculate whether the lag phase was mainly due to a die-off of flagellate cells in the inoculum or due to a prolonged phase of inactivity of the inoculated cells. In order to investigate the extent of flagellate mortality during the early phases of the experiment, we set up an experiment with very large inocula. This enabled us to track possible decreases in cell numbers after inoculation.

Five millilitres of TSB medium in sterile plastic tubes was inoculated with 10 μL of the two precultures (R. fascians and S. witflariensis). The tubes were placed on a shaker in a 15 °C climate chamber. The bacterial cultures were allowed to reach the stationary phase (after 4 days). The bacteria were harvested by centrifugation in 10-mL centrifuge tubes (4000 g, 30 min). In part of the tubes, the bacteria were subjected to a single washing (with 5 mL sterile AS) and centrifugation cycle, after which the bacteria were resuspended in 5 mL sterile AS. In tubes where we wanted to obtain unwashed bacterial cultures, the bacteria were simply resuspended in the supernatant.

We used the following treatments: washed bacterial cells (WB) of R. fascians grown in 0.3 and 1 g TSB L−1, unwashed bacterial culture of R. fascians (UC) grown in 0.3 and 1 g TSB L−1, washed cells of S. witflariensis (WB) grown in 1 g TSB L−1 and finally dilution with sterile AS. All treatments were prepared by adding 4 mL of the cell suspensions or AS to triplicate cell culture flasks. An inoculum of c. 20 000 cells of the heterotrophic flagellate Cercomonas sp. from an actively growing axenic culture was added to each flask, giving rise to an estimated starting concentration of 5000 cells mL−1.

Flagellate counting procedure

The growth of the flagellate cultures was followed via direct counting in the cell culture flasks using an inverted microscope at × 200 magnification. We counted the number of cells in 21 (Experiments 1, 2 and 3A) or 42 (Experiments 3B, 3C and 4) counting grids of variable size (0.0025–0.25 mm2) in each flask at each counting session.

Statistical analyses

All regressions were performed in sigmaplot 2001 (Systat Software Inc.), while two-way anovas and multiple comparison tests were performed using sas enterprise guide V. 4.1 (SAS Institute). CFU data were tested for the two bacteria (R. fascians and S. witflariensis) individually. The CFU data from Day 21 were transformed (square root) to obtain equal variances. Levene's test (P>0.05) was used to check for equal variances.

Calculations of Cercomonas sp. growth parameters

Growth rate

Following Koch & Ekelund (2005), we assumed that the natural logarithm (ln) of the flagellate abundance was a sigmoid function of time. We therefore fitted the natural logarithm of the population data to a three-parameter sigmoid function of the following form:

  • image(1)

As described by Koch & Ekelund (2005), the specific growth rate (μ) is obtained by differentiating Eqn. (1) in the vertical inflection point (t0) of the function:

  • image(2)
Peak abundance

The peak abundances are based on the three-parameter sigmoid curve [Eqn. (1)] fitted to untransformed population data (see Fig. 1).

image

Figure 1.  Growth of Cercomonas sp. when feeding on different combinations of the two bacteria (Experiment 1). The two food bacteria are supplied in seven different cell ratios (based on direct microscopic counts) from 100%Rhodococcus fascians (100/0) to 100%Sphingopyxis witflariensis (0/100). The growth data are fitted to a 3-parameter sigmoid curve [Eqn. (1)]. The values represent means of three replicates.

Download figure to PowerPoint

In Experiment 1, we controlled the cell number of the two bacteria. The different treatments (different cell ratios of the bacteria) in this experiment therefore do not necessarily represent the same amount of food for the flagellates. We therefore calculated the yield of Cercomonas sp. cells per cell volume of the bacteria in the different treatments.

Lag phase

Baranyi & Pin (1999) use a definition of the lag phase as the time from the beginning of the experiment until the tangent drawn to the exponential phase of the growth curve intercepts with the inoculum level. Thus, by assuming exponential growth of the flagellate cultures from t=0 to t=t0, we can calculate the length of the lag phase (tlag) as

  • image(3)

If tlag>0, the population growth has been slower than expected from the growth rate alone. In this situation there is a proper lag phase. A tlag<0 could be due to an underestimation of the inoculum size or the growth rate.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion and perspectives
  8. Acknowledgements
  9. References

Experiment 1: Growth on washed bacteria in different ratios

Growth parameters

The growth pattern of Cercomonas sp. with different ratios of washed cells of R. fascians to S. witflariensis used as food for the flagellates was similar but the growth rates and peak abundances were different (Fig. 1). The growth rates showed significant differences between the two pure suspensions and some of the other combinations (Fig. 2a). It is noteworthy that we observed a significant reduction in the growth rate of Cercomonas sp. when adding only 10%R. fascians cells, while an addition of 30%S. witflariensis cells did not lead to a significant increase in the growth rate compared with the treatment with 100%R. fascians. The number of flagellates produced per biovolume of bacteria present at the onset of the experiment was higher when Cercomonas sp. was fed with washed cells of R. fascians compared with washed cells of S. witflariensis (Fig. 2b).

image

Figure 2.  Growth data from Experiment 1. The growth rate (a) of Cercomonas sp. in the different treatments. Different letters above the bars represent significantly different values (Tukey's multiple comparisons test, P<0.05). Flagellate yield (b) calculated by dividing the peak abundance of Cercomonas sp. by the corresponding total biovolume of bacteria at the onset of the experiment. The values represent means of three replicates (error bars represent ± SE).

Download figure to PowerPoint

CFU data

The different colony morphology of the two bacteria enabled us to follow changes in the numbers of the two bacteria. At Day 11 (Fig. 3a), there was a significant effect of grazing on the CFUs of S. witflariensis (two-way anova: P=0.0082). This effect was only significant in the treatment with S. witflariensis (Tukey's multiple comparisons test: P<0.05), whereas none of the two bacteria were affected by grazing in the 50/50 treatment (Tukey's multiple comparisons test: P>0.05). At Day 21 (Fig. 3b), there was a significant effect of grazing on the CFUs of R. fascians (two-way anova: P<0.0001). This effect was significant for all the bacterial ratios tested (Tukey's multiple comparisons test: P<0.05). No effect of grazing was detectable on the CFUs of S. witflariensis at this day (two-way anova: P>0.05).

image

Figure 3.  CFU of the two food bacteria (Rhodococcus fascians and Sphingopyxis witflariensis) from flasks with (+) or without (–) flagellates at day 11 (a) and day 21 (b). At Day 11, CFU counts were performed on the treatments 50/50 and 0/100 (R. fascians/S. witflariensis) and at Day 21 they were performed on flasks from the treatments: 100/0; 70/30; 50/50; 30/70 and 0/100. The values represent means of three replicates (error bars represent ±SE).

Download figure to PowerPoint

Experiment 2: Growth on washed bacteria with fresh TSB medium

The general growth pattern in this experiment (Fig. 4a) was similar to the one observed in Experiment 1 (Fig. 1): there was a fast initial growth on washed cells of S. witflariensis compared with the bacterial mixture and washed R. fascians cells (Fig. 4a). The specific growth rate (Fig. 4b) in flasks supplied with additional TSB during the experiment was higher compared with flasks given additional washed bacteria (two-way anova: P<0.0001). In the pair-wise comparisons, this was only significant for S. witflariensis as food (Tukey's multiple comparisons test: P<0.05).

image

Figure 4.  Growth of Cercomonas sp. in three different types of bacterial suspensions: Rhodococcus fascians only (R), Sphingopyxis witflariensis only (S) or a 50/50 mixture (M) of the two (Experiment 2). The cultures were supplemented with either fresh TSB medium (FM) or additional washed bacteria (WB) at Days 5, 8 and 23. Number of Cercomonas sp. cells as a function of time (a), growth rate of Cercomonas sp. (b) and flagellate peak abundance (c). The values represent means of three replicates (error bars represent ±SE).

Download figure to PowerPoint

The peak abundance of Cercomonas sp. was again higher when fed with R. fascians compared with S. witflariensis while the peak abundance with the 50/50 mixture was intermediary (Fig. 4c). There was an overall positive effect on the flagellate peak abundance of adding fresh TSB during the experiment compared with adding additional washed bacteria (two-way anova: P<0.0001). This effect was most pronounced with S. witflariensis, less with the mixture and not significant with R. fascians (Tukey's multiple comparisons test: P<0.05). This resulted in a significant interaction between addition during the experiment (TSB or washed bacteria) and type of bacteria (R. fascians, 50/50 mixture or S. witflariensis) (two-way anova: P=0.0006).

The positive effect on flagellate growth parameters of adding fresh bacterial medium compared with washed cells of R. fascians (Fig. 4) clearly indicated that the ability of flagellates to obtain high densities on washed Rhodococus cells after a long incubation time (Experiment 1, Fig. 1) was not due to a change in the properties of the washed bacterial cells caused by starvation.

Experiment 3: Growth on unwashed bacteria

Growth rates

The specific growth rate of Cercomonas sp. was again much lower when fed with R. fascians compared with when it was fed with S. witflariensis (Fig. 5a). When unwashed cultures of S. witflariensis were used as food, growth was observed over the entire TSB concentration spectrum, and the growth rates tended towards a unimodal relationship with the TSB concentration on this food bacterium. Growth was not observed above TSB concentrations of 0.2 g TSB L−1 when the flagellates were inoculated into unwashed cultures of R. fascians. For one of the TSB concentrations (0.06 g L−1), we found an unusually high growth rate (this point is marked with an arrow). We have no explanation for this outlying point and it has no equivalent in the two additional experiments (data not shown). Otherwise, these experiments showed similar relationships between TSB concentration and the basic growth parameters. When averaging for all TSB concentrations, the mean growth rate on the mixture (0.75 day−1) was above the mean growth rate on R. fascians (0.23 day−1) and below the value on S. witflariensis (1.55 day−1). Furthermore, the value for the mixture was at the same level as the grand mean for R. fascians and S. witflariensis (0.84 day−1).

image

Figure 5.  Growth rate (a) and flagellate peak abundance (b) as a function of the TSB concentration (Experiment 3). Cercomonas sp. was grown with unwashed culture of either Rhodococcus fascians, Sphingopyxis witflariensis or a 50/50 mixture of the two as food. The values represent means of three replicates (error bars represent ±SE). The two arrows represent a TSB concentration where the growth on R. fascians deviated from the general trend on this bacterium.

Download figure to PowerPoint

Peak abundance of Cercomonas sp.

Overall, there was a unimodal relationship between TSB concentration and peak abundance on the different bacterial food (Fig. 5b). When supplied with unwashed culture of the two bacteria, the situation was somewhat different than that with washed cells. At the lower food concentrations, there was a small difference in the peak abundance of the flagellates between the three bacterial food types (R. fascians, S. witflariensis and the 50/50 mixture), the highest peak abundance being obtained with S. witflariensis as food. At higher bacterial densities, however, this difference became more pronounced. The Cercomonas sp. population reached a higher maximum cell density on S. witflariensis, and did so at a higher concentration of TSB (0.6 g TSB L−1), compared with the results with R. fascians only. On R. fascians, the flagellates did not grow at concentrations above 0.2 g TSB L−1. With the mixture of the two bacteria, the flagellates grow at all the concentrations used (0.03–1.2 g TSB L−1) but only reached very limited peak abundances above 0.3 g TSB L−1. At the highest concentration, the flagellates were only able to maintain a detectable population transiently (data not shown).

Lag phase

The lag phase was generally longer when Cercomonas sp. was fed with an unwashed culture of R. fascians compared with S. witflariensis and the mixture of the two bacteria (Fig. 6). When Cercomonas sp. was supplied with R. fascians, there was a positive linear relationship between TSB concentration and the length of the lag phase (Fig. 6a) (P=0.0019). No such relationship was seen with S. witflariensis or the mixture (Fig. 6b and c).

image

Figure 6.  Lag time (tlag) of Cercomonas sp. (see text) as a function of the TSB concentration (Experiment 3). For cultures with Rhodococcus fascians (a), there was a positive linear relationship between the two variables (linear regression: P=0.0019). No such relationship was found for cultures with the mixture of the two food bacteria (b) and cultures with Sphingopyxis witflariensis (c).

Download figure to PowerPoint

Experiment 4: Large inoculum size

In this experiment the number of flagellates added to the flasks was higher than the level of detection. This made it possible to track any reductions in the flagellate abundance. In Fig. 7, the flagellate abundances are plotted as a function of time. For practical reasons, we only counted the number of flagellates at Day 0 in the AS-diluted flasks. In the following examination of the data, we assume that the density of flagellates in all flasks had started at the same level as in the AS-diluted flasks. It was observed that starvation (dilution with AS) made the flagellates die off at a constant rate over the time period examined (linear regression: P=0.0021).

image

Figure 7.  Change in density of Cercomonas sp. cells after inoculation into different bacterial suspensions or into a salt solution without bacteria. The bacteria were grown at two different TSB concentrations and supplied as untreated cultures or as washed cells. Treatments: food bacterium [Sphingopyxis witflariensis (S) or Rhodococcus fascians (R)], TSB concentration (1.0 or 0.3 g TSB L−1) and bacterial treatment [washed bacteria (WB) or untreated bacterial culture (UC)]. In the control treatment (AS), the flagellates were added to a salt solution (amoebae saline) without bacterial cells.

Download figure to PowerPoint

The flagellates died off faster during the first 24 h when they were supplied with unwashed R. fascians culture and they continued to die off until they declined below the detection level in two out of the three flasks with 1 g TSB L−1 culture. In the third of these flasks, growth was detected after Day 3. In the treatment with 0.3 g TSB L−1 culture, we observed growth of the flagellates after Day 2. When washed cells of R. fascians were used as food, we only observed a small initial decline (of the same magnitude as in the AS-diluted flasks) in flagellate numbers. Here, the growth rate of the flagellates was enhanced by a higher density of R. fascians cells (1.0 vs. 0.3 g TSB L−1 treatments). As observed earlier, the growth rate of Cercomonas sp., when supplied with washed cells of S. witflariensis, was considerably higher than when they were supplied with washed R. fascians cells.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion and perspectives
  8. Acknowledgements
  9. References

Basic flagellate growth parameters depended on the food bacterium offered

The protozoa have a very important role in soil as grazers of bacteria (Clarholm, 1981). When conditions become suitable for bacterial growth, the protozoa are the first grazers to increase in numbers due to their high growth rates (Clarholm, 1981; Ekelund, 1996; Rønn et al., 1996). Active protozoa have very limited mobility in soil (Griffiths & Caul, 1993) and so the fast response of protozoa to rapid bursts in bacterial production is dependent on high growth rates of the protozoa (Vestergaard et al., 2001).

The maximum specific growth rate of Cercomonas sp. was higher on S. witflariensis than on R. fascians, irrespective of the way in which the bacteria were presented to the flagellates: washed bacteria with or without fresh TSB medium or unwashed bacterial culture. This is in agreement with previous studies showing poor growth of single species of amoebae and flagellates on actinobacteria (Weekers et al., 1993; Mohapatra & Fukami, 2005). Similarly, Rønn et al. (2001) found a complete absence of growth of populations of indigenous soil protozoa when amended with the actinobacterium Mycobacterium chlorophenolicum compared with an amendment with Pseudomonas chlororaphis.

The slow growth rate of Cercomonas sp. on R. fascians was possibly related directly to the bacterial cells because we observe the same effect on washed (Figs 2a and 4b) and unwashed cells (Fig. 5a). A slower degradation of the bacterial cells would result in a slower growth rate of the flagellates. Slower degradation of low-G/C Gram-positive bacteria compared with Gram-negative bacteria has been shown in vacuoles of flagellates and ciliates (Gonzalez et al., 1990; Iriberri et al., 1994). This has been related to the thicker cell wall of the Gram-positive types (Iriberri et al., 1994).

In spite of the higher growth rate of Cercomonas when supplied with S. witflariensis compared with when it is supplied with R. fascians, it appeared that the cell yield of Cercomonas sp. was higher on washed cells of R. fascians compared with S. witflariensis (Figs 2b and 4c). Sherr et al. (1983) observed an inverse relationship between the growth rate of the flagellate Monas sp. and its cell yield when fed with four different bacteria (three proteobacteria and one unknown), and so the two growth parameters are not necessarily positively correlated.

The observed lower peak abundance of Cercomonas growing on S. witflariensis compared with R. fascians could be due to food limitation, or a deficiency of some essential compound, when the flagellates were fed with S. witflariensis. In Experiment 2, we observed that an addition of fresh TSB had a larger effect on peak abundance and growth rate on Sphingopyxis than on Rhodococcus (Fig. 4b and c). This suggests that the starved Sphingopyxis cells used the fresh TSB for synthesis of various cell constituents that made them more suitable as food for the flagellates, whereas the food quality of Rhodococcus cells was only slightly affected by addition of a fresh substrate. We cannot rule out that part of the TSB added during the experiment could be utilized directly by the flagellates. However, this is probably of minor importance because the starved bacterial cells would be much more efficient in utilizing the substrate than the flagellates due to the higher cell density and their lower surface/volume ratio (Fenchel, 1987).

There was a marked difference in the growth patterns obtained with washed bacterial cells and with unwashed cultures. When flagellates were supplied with washed cells, we found that the highest peak abundances arose with R. fascians as the food source. However, when they were supplied with unwashed bacterial cultures, we found the highest obtainable peak abundances on S. witflariensis. This effect was most pronounced at relatively high to medium concentrations where Cercomonas sp. appeared to be inhibited by the presence of R. fascians, either separately or in mixture (Fig. 5b). The conclusion is that R. fascians was able to support high abundances of Cercomonas sp. but only if the bacteria were presented as washed cells. This suggests that there was some soluble inhibitory substances in the spent growth media from the R. fascians cultures.

Does Cercomonas sp. graze selectively ?

Jezbera et al. (2005) observed a negative selection of actinobacteria by a natural aquatic protozoan community. The observations by Rønn et al. (2002) that actinobacteria are favoured in soil microcosms when protozoa are present could also be explained by selective feeding of protozoa. Therefore, we hypothesized that Cercomonas sp. would avoid engulfing R. fascians and thus show selectivity towards S. witflariensis. This idea is not supported by our data. The CFU numbers indicated a statistically significant reduction in the numbers of R. fascians caused by flagellate grazing at Day 21 (Fig. 3b). Although the CFU data may have been biased by cryptic growth of the bacteria during the experiment (Ekelund, 1996; Snyder & Hoch, 1996), these results indicate that, if the flagellates were selecting between the two bacteria, they preferred R. fascians. The positive correlation between the proportion of R. fascians and flagellate yield (Fig. 2b) also indicates that Cercomonas sp. was consuming R. fascians. Furthermore, the more than proportionate negative effect of adding R. fascians on the growth rate in Experiment 1 (Fig. 2a), and the fact that the growth rate of the flagellates when feeding on the 50/50 mixture in Experiment 3 is more or less an average of the growth rate when feeding on pure cultures of the two bacteria (Fig. 5b), supports the idea that the flagellates are not avoiding R. fascians.

Although we observed the highest growth rates in the treatment with Sphingopyxis alone (Experiment 1), the grazing pressure exerted by the flagellates during the whole experiment was expected to be much higher in flasks with Rhodococcus cells as the only food source because the peak abundance of the flagellates was much higher when fed with Rhodococcus compared with Sphingopyxis (Fig. 1). Therefore, given the fact that the flagellates were indeed eating Rhodococcus cells, it is not surprising that we observed a much larger reduction in bacterial cell numbers in the treatment with R. fascians cells as food compared with the treatment with S. witflariensis as the sole food source. These results demonstrate that the potential abundance of flagellates – and thereby their potential importance as grazers – when growing on a certain combination of bacteria cannot be fully described by the growth rate. If we observe a high growth rate, we cannot necessarily conclude that the flagellates will have a substantial effect as grazers because a high growth rate does not necessarily lead to a high abundance.

Does bacterial food quality depend on the TSB concentration?

The relationship between the density of food and growth rates of protozoa has often been described as a form of saturation kinetics (Taylor, 1978; Fenchel, 1987; Ekelund, 1996). But protozoa can be inhibited in their growth at very high food densities (Ekelund, 1996), which is probably related to a build-up of toxic bacterial products and a deprivation of oxygen (Singh, 1941). Rønn et al. (1995) observed a negative correlation between the number of soil protozoa in MPN numerations and the TSB concentration (0.03–3 g TSB L−1). The build-up of various extracellular metabolites is potentially very important in soil, where most of the bacterial growth (and hence flagellate growth) is situated in hotspots (Coleman, 1994; Ekelund & Rønn, 1994) as for instance the rhizosphere (Bonkowski, 2004) or around decomposing organic matter (Christensen et al., 1992; Rønn et al., 1996).

In this study, we found clear effects of TSB concentration on flagellate growth (Figs 5a, b and 6a). The effect was very evident for R. fascians, which did not support growth at high TSB concentrations, whereas there was only a minor negative effect of TSB concentration on flagellate growth on S. witflariensis. Our original hypothesis, was that the low food quality of R. fascians would be related to the Gram-positive natures of the cell wall. However, if the poor food quality is merely a consequence of the thick cell wall, we would expect any limitation in the growth of Cercomonas sp. due to concentration-dependent factors (e.g. waste products and oxygen deprivation) to be similar for the two bacteria. Likewise, we would expect the effect of washing the bacteria on the growth of Cercomonas sp. to be the same for the two bacteria. Because the negative effects of TSB concentration were much larger for R. fascians than for S. witflariensis, we conclude that it is unlikely that differences in the composition of the cell wall were the only reason for the observed difference in food quality between the two bacteria; there are probably some additional inhibitory compounds present in the growth medium from the Rhodococcus cultures that are not present in the S. witflariensis cultures. The small inhibiting effect of high TSB concentrations on the growth of Cercomonas sp. when S. witflariensis was the only food source (Fig. 5a and b) might, to a large extent, be coupled to a deprivation of oxygen in the flasks.

The longer lag phase (tlag) when flagellates are grown with R. fascians as food than when they are grown with S. witflariensis (Fig. 6) could be related to a slow change in flagellate physiology to accommodate a possible toxicity or low digestibility of the Rhodococcus cells. Alternatively, it could be related to a more drastic reduction in flagellate cell numbers following inoculation into the newly established cultures. In many growth experiments, it can be difficult to discriminate between these two possibilities; it is often not known how large a part of the inoculated cells survives the inoculation procedure in a physiological state that allows them to resume growth. In this study, we attempted to evaluate this by adding a larger inoculum and follow the initial change in cell number. The results of this experiment (Fig. 7) clearly suggest that flagellate cells die off when added to unwashed Rhodococcus cultures; only a minor proportion of the cells survive and are able to grow (Fig. 7). This indicates that part of the reason for the longer lag phase of the flagellates when added to Rhodococcus cultures is that the actual initial number of cells that are able to start growth is lower than when they are added to Sphingopyxis cultures. This does not rule out the possibility that a slow physiological adaptation to feeding on Rhodococcus cells is also involved in the delayed growth response of the flagellates when feeding on Rhodococcus.

These results also suggest that the lack of growth with Rhodococcus as food at the highest TSB concentrations in Experiment 3 was partly a consequence of the relatively small inoculum of flagellate cells, because we detected growth in one out of three flasks with unwashed R. fascians culture at a concentration of 1 g TSB L−1 in Experiment 4 (Fig. 7).

A possible explanation for the marked negative effect of high TSB concentrations in treatments with R. fascians could be that the R. fascians cells changed composition when grown at high concentrations of medium, but in Experiment 4 we observed that the washed cells of R. fascians were not less suitable as food for Cercomonas sp. when grown at relatively high concentrations of medium (1 g TSB L−1), while unwashed cells at these densities were seen to be toxic for the flagellates (Fig. 7). As noted above, Cercomonas sp. was only able to build up dense populations of cells on R. fascians if these had been washed before the experiment. Hence, washing the bacterial cells had profound effects on the food quality of R. fascians for the flagellate. These results indicate that R. fascians is able to produce one or more extracellular metabolites that are harmful for the flagellates – and that more of these harmful metabolites is produced with increased TSB concentration.

Conclusion and perspectives

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion and perspectives
  8. Acknowledgements
  9. References

The actinobacterium R. fascians was shown to be of lower food quality for an axenic culture of the flagellate Cercomonas sp. than the alphaproteobacterium S. witflariensis. The poor food quality of R. fascians was partly due to factors that were independent of the medium concentration, possibly related to the thick actinobacterial cell wall, partly due to a concentration-dependent factor, which we hypothesize was one or more harmful extracellular metabolites produced by the bacterium. However, as opposed to Bjørnlund et al. (2006), we could observe a growth of Cercomonas sp. on the actinobacterium, but it was slow and often very delayed due to the toxicity of R. fascians at high medium concentrations. The flagellates were not avoiding the actinobacterium R. fascians when supplied with the two food bacteria in combination. One might argue that there is a discrepancy between this result and earlier investigations showing that protozoa in soil are able to favour Gram-positive bacteria (Griffiths et al., 1999) or specifically actinobacteria (Rønn et al., 2002). This could be due to predator-specific differences in selectivity among protozoa (Pernthaler et al., 2001). Another possible explanation lies in the heterogeneity of the soil environment. We propose that this heterogeneity can lead to a mosaic of microhabitats in soil with a predominance of different bacteria. In patches with a predominance of actinobacteria – of relatively low food quality – we would expect a more limited (and delayed) protozoan growth, thus resulting in a relatively low grazing pressure on the actinobacteria in the short-time perspective.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion and perspectives
  8. Acknowledgements
  9. References

We thank Karin Westberg, Section for Microbiology, for help with technical issues, Carsten Suhr Jacobsen, Geological Survey of Denmark and Greenland, for comments on an earlier draft of the manuscript and staff and students at the Section for Terrestrial Ecology for good discussions, technical help and a nice working environment.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion and perspectives
  8. Acknowledgements
  9. References
  • Andersen KS & Winding A (2004) Non-target effects of bacterial biological control agents on soil Protozoa. Biol Fertil Soils 40: 230236.
  • Baranyi J & Pin C (1999) Estimating bacterial growth parameters by means of detection times. Appl Environ Microbiol 65: 732736.
  • Bjørnlund L, Mørk S, Vestergård M & Rønn R (2006) Trophic interactions between rhizosphere bacteria and bacterial feeders influenced by phosphate and aphids in barley. Biol Fertil Soils 43: 111.
  • Bonkowski M (2004) Protozoa and plant growth: the microbial loop in soil revisited. New Phytol 162: 617631.
  • Chang SL (1960) Growth of small free-living amoebae in various bacterial and in bacteria-free cultures. Can J Microbiol 6: 397405.
  • Christensen S & Bonde GJ (1985) Seasonal variation in numbers and activity of denitrifying bacteria in soil. Taxonomy and physiological groups among isolates. Tidsskr planteavl 89: 367372.
  • Christensen S, Griffiths BS, Ekelund F & Rønn R (1992) Huge increase in bacterivores on freshly killed barley roots. FEMS Microbiol Ecol 86: 303310.
  • Chrzanowski TH & Šimek K (1990) Prey-size selection by freshwater flagellated protozoa. Limnol Oceanogr 35: 14291436.
  • Clarholm M (1981) Protozoan grazing of bacteria in soil – impacts and importance. Microb Ecol 7: 343350.
  • Coleman DC (1994) The microbial loop concept as used in terrestrial soil ecology studies. Microbial Ecology 28: 245250.
  • DeRuiter PC, Moore JC, Zwart KB et al. (1993) Simulation of nitrogen mineralization in the below-ground food webs of two winter wheat fields. J Appl Ecol 30: 95106.
  • Ekelund F (1996) Growth kinetics of five common heterotrophic soil flagellates. Eur J Soil Biol 32: 1524.
  • Ekelund F & Rønn R (1994) Notes on protozoa in agricultural soil with emphasis on heterotrophic flagellates and naked amoebae and their ecology. FEMS Microbiol Rev 15: 321353.
  • Ekelund F, Rønn R & Griffiths BS (2001) Quantitative estimation of flagellate community structure and diversity in soil samples. Protist 152: 301314.
  • Fenchel T (1987) Ecology of Protozoa: The Biology of Free-Living Phagotrophic Protists. Science Tech Publishers, Madison, WI.
  • Gonzalez JM (1996) Efficient size-selective bacterivory by phagotrophic nanoflagellates in aquatic ecosystems. Mar Biol 126: 785789.
  • Gonzalez JM, Iriberri J, Egea L & Barcina I (1990) Differential rates of digestion of bacteria by freshwater and marine phagotrophic protozoa. Appl Environ Microbiol 56: 18511857.
  • Griffiths BS (1989) Enhanced nitrification in the prescence of bacteriophagous protozoa. Soil Biol Biochem 21: 10451051.
  • Griffiths BS & Caul S (1993) Migration of bacterial-feeding nematodes, but not protozoa, to decomposing grass residues. Biol Fertil Soils 15: 201207.
  • Griffiths BS, Bonkowski M, Dobson G & Caul S (1999) Changes in soil microbial community structure in the presence of microbial-feeding nematodes and protozoa. Pedobiologia 43: 297304.
  • Groscop JA & Brent MM (1964) The effects of selected strains of pigmented microorganisms on small free-living amoebae. Can J Microbiol 10: 579584.
  • Hahn MW & Höfle MG (2001) Grazing of protozoa and its effect on populations of aquatic bacteria. FEMS Microbiol Ecol 35: 113212.
  • Hunt HW, Coleman DC, Ingham ER, Ingham RE, Elliot ET, Moore JC, Rose SL, Reid CPP & Morley CR (1987) The detrital food web in a shortgrass prairie. Biol Fertil Soils 3: 5768.
  • Iriberri J, Azúa I, Labirua-Iturburu A, Artolozaga I & Barcina I (1994) Differential elimination of enteric bacteria by protists in a freshwater system. J Appl Bacteriol 77: 476483.
  • Jezbera J, Horòák K & Šimek K (2005) Food selection by bacterivorous protists: insight from the analysis of the food vacuole content by means of fluorescence in situ hybridization. FEMS Microbiol Ecol 52: 351363.
  • Jousset A, Lara E, Wall LG & Valverde C (2006) Secondary metabolites help biocontrol strain Pseudomonas fluorescens CHA0 to escape protozoan grazing. Appl Environ Microbiol 72: 70837090.
  • Jürgens K & DeMott WR (1995) Behavioral flexibility in prey selection by bacterivorous nanoflagellates. Limnol Oceanogr 40: 15031507.
  • Kinner NE, Harvey RW, Blakeslee K, Novarino G & Meeker LD (1998) Size-selective predation on groundwater bacteria by nanoflagellates in an organic-contaminated aquifer. Appl Environ Microbiol 64: 618625.
  • Koch TA & Ekelund F (2005) Strains of the heterotrophic flagellate Bodo designis from different environments vary considerably with respect to salinity preference and SSU rRNA gene composition. Protist 156: 97112.
  • Matz C & Jürgens K (2003) Interaction of nutrient limitation and protozoan grazing determines the phenotypic structure of a bacterial community. Microb Ecol 45: 384398.
  • Matz C, Deines P & Jürgens K (2002) Phenotypic variation in Pseudomonas sp. CM10 determines microcolony formation and survival under protozoan grazing. FEMS Microbiol Ecol 39: 5765.
  • Matz C, Deines P, Boenick J, Arndt H, Eberl L, Kjelleberg S & Jürgens K (2004) Impact of violacein-producing bacteria on survival and feeding of bacterivorous nanoflagellates. Appl Environ Microbiol 70: 15931599.
  • Mohapatra BR & Fukami K (2005) Effect of different bacterial species on the growth kinetics of the heterotrophic nanoflagellate Jakoba libera. Basic Appl Ecol 6: 6773.
  • Page FC (1988) A New Key to Freshwater and Soil Gymnamoebae. Freshwater Biological Association.
  • Pernthaler J, Posch T, Šimek K, Vrba J, Pernthaler A, Glöckner FO, Nübel U, Psenner U & Amann R (2001) Predator-specific enrichment of actinobacteria from a Cosmopolitan Freshwater Clade in mixed continuous culture. Appl Environ Microbiol 67: 21452155.
  • Posch T, Šimek K, Vrba J, Pernthaler J, Nedoma J, Sattler B, Sonntag B & Psenner R (1999) Predator-induced changes of bacterial size-structure and productivity studied on an experimental microbial community. Aquat Microb Ecol 18: 235246.
  • Rønn R, Ekelund F & Christensen S (1995) Optimizing soil extract and broth media for MPN-enumeration of naked amoebae and heterotrophic flagellates in soil. Pedobiologia 39: 1019.
  • Rønn R, Griffiths BS, Ekelund F & Christensen S (1996) Spatial distribution and successional pattern of microbial activity and micro-faunal populations on decomposing barley roots. J Appl Ecol 33: 662672.
  • Rønn R, Grunert J & Ekelund F (2001) Protozoan response to addition of the bacteria Mycobacterium chlorophenolicum and Pseudomonas chlororaphis. Biol Fertil Soils 33: 126131.
  • Rønn R, McCaig AE, Griffiths BS & Prosser JI (2002) Impact of protozoan grazing on bacterial community structure in soil microcosms. Appl Environ Microbiol 68: 60946105.
  • Sherr BF, Sherr EB & Berman T (1983) Grazing, growth, and ammonium excretion rates of a heterotrophic microflagellate fed with four species of bacteria. Appl Environ Microbiol 45: 11961201.
  • Šimek K, Horňák K, Jezbera J, Mašín M, Nedoma J, Gasol JM & Schauer M (2005) Influence of top-down and bottom-up manipulations on the R-BT065 subcluster of β-proteobacteria, an abundant group in bacterioplankton of a freshwater reservoir. Appl Environ Microbiol 71: 23812390.
  • Singh BN (1941) The influence of different bacterial food supplies on the rate of reproduction in Colpoda steinii, and the factors influencing encystations. Ann Appl Biol 28: 6573.
  • Singh BN (1945) The selection of bacterial food by soil amoebae, and the toxic effects of bacterial pigments and other products on soil protozoa. Br J Exp Pathol 26: 316325.
  • Snyder RA & Hoch MP (1996) Consequences of protist-stimulated bacterial production for estimating protist growth efficiencies. Hydrobiologia 41: 113123.
  • Taylor WD (1978) Growth responses of ciliate protozoa to the abundance of their bacterial prey. Microb Ecol 4: 207214.
  • Verhagen FJM, Duyts H & Laanbroek HJ (1993) Effect of grazing by flagellates on competition for ammonium between nitrifying and heterotrophic bacteria in soil columns. Appl Environ Microbiol 59: 20992106.
  • Vestergård M, Bjørnlund L & Christensen S (2004) Aphid effects on rhizosphere microorganisms and microfauna depend more on barley growth phase than on soil fertilization. Oecologia 141: 8493.
  • Vestergaard P, Rønn R & Christensen S (2001) Reduced particle size of plant material does not stimulate decomposition, but affects the microbivorous microfauna. Soil Biol Biochem 33: 18051810.
  • Weekers PHH, Bodelier PLE, Wijen JPH & Vogels GD (1993) Effects of grazing by the free-living soil amoebae Acanthamoeba castellanii, Acanthamoeba polyphaga, and Hartmanella vermiformis on various bacteria. Appl Environ Microbiol 59: 23172319.
  • Wootton EC, Zubkov MV, Jones DH, Jones RH, Martel CM, Thornton CA & Robert EC (2007) Biochemical prey recognition by planktonic protozoa. Environ Microbiol 9: 216222.