Survival and Spread of GMMs in Soil columns Field lysimeters and Field Plots

In the greenhouse experiment, a continuous decline of the inoculated cells was observed over a period of 85 weeks (Fig. 9.3). Strain S. melilotiL33 declined to 9.0 x 104 cfu g-1 soil within 24 weeks and to 2.8 x 103 cfu g-1 within 85 weeks in the upper 25 cm of the soil columns.23 Decline rates for S. meliloti L1 were not significantly different, indicating that in the presence of its host plant, alfalfa, and under greenhouse conditions, the recA-mutation did not affect the environmental fitness of S. meliloti. In order to confirm that the decrease of the GMM was due to its elimination from the soil habitat, and not just a decrease of culturable cells (for more details of this phenomenon, see Chapter 1), we additionally monitored the presence of the luc marker gene in DNA directly extracted from soil by PCR. Techniques to obtain DNA from soil suitable for PCR detection are described in Chapter 3. Using carefully selected primers, we could differentiate between product sizes obtained from strain L1 and L33 (for positions of primers, see Fig. 9.1). By this means we were able to show that no cross-contamination of inoculated GMM cells occurred between the lysimeters (Fig. 9.4). Additionally, following GMM inoculation, a decrease in PCR product yields was observed. This decrease was probably caused by a decline of the template copy numbers in soil DNA.

A similar decrease of inoculated cells was detected after field lysimeter inoculation: after 24 weeks, 2.0 x 104 cfu g-1 soil were detected for strains L1 and L33, respectively. In contrast to the greenhouse experiment, however, the titer did not decrease any further, until growth of alfalfa was stopped after 80 weeks.24 A similar population decline as observed in the field lysimeters, was detected after inoculation of field plots. Populations of both strains dropped from 106 to below 104 within 14 weeks, then increased to 2 x 104 cfu g-1, 24 weeks after inoculation. Until today, four years after the field release, the titers of L33 and L1

Fig. 9.3. Survival of S. meliloti strains in the upper 25 cm in soil columns (greenhouse). Strain L1 (▲) and strain L33 (•). Reprinted with permission from: Schwieger F, Willke B, Munch JC et al. Biol Fertil Soils 1997; 25:340-348.

Fig. 9.4. PCR mediated detection of the recombinant /uc-marker gene in DNA, directly extracted from soil. PCR products of strain L33 (415 base pairs; lanes 3, 5, 7, 9, 11, 13, 16) and L1 (1011 base pairs; lanes 2, 4, 6, 8, 10, 12 and 15) are clearly distinguishable. DNA was extracted from soil columns (0-25 cm depth) after a day (lanes 2, 3), 2.1 weeks (4, 5), 4.1 weeks (6, 7), 8.1 weeks (8, 9), 16.1 weeks (10, 11), and 24.1 weeks (12,13). Other lanes show size standards and controls. Reprinted with permission from: Schwieger F, Willke B, Munch JC et al. Biol Fertil Soils 1997; 25:340-348.

Fig. 9.4. PCR mediated detection of the recombinant /uc-marker gene in DNA, directly extracted from soil. PCR products of strain L33 (415 base pairs; lanes 3, 5, 7, 9, 11, 13, 16) and L1 (1011 base pairs; lanes 2, 4, 6, 8, 10, 12 and 15) are clearly distinguishable. DNA was extracted from soil columns (0-25 cm depth) after a day (lanes 2, 3), 2.1 weeks (4, 5), 4.1 weeks (6, 7), 8.1 weeks (8, 9), 16.1 weeks (10, 11), and 24.1 weeks (12,13). Other lanes show size standards and controls. Reprinted with permission from: Schwieger F, Willke B, Munch JC et al. Biol Fertil Soils 1997; 25:340-348.

remained in the range between 2 x 103 and 7 x 104 cfu g-1 soil, with a seasonal impact on the population sizes. Interestingly, in the first two years, the populations of L1 were below those of L33 in Fall and Winter but above L33 in Spring, as observed in the two following years.25 This suggested that there was an ecological significance of the recA mutation. However, in the third year, this phenomenon was not significant. Even though it is extremely difficult to determine the reason for significant differences between recA- and recA+, the majority of sampling dates did not yield such differences and, thus, we can conclude that the recA gene was not crucial for S. meliloti to successfully colonize the field plots with alfalfa. Due to the high persistence of both strains, L1 and L33, we can also conclude that the luciferase marker gene did not interfere with the environmental fitness of S. meliloti. This clearly supports the assumption that in contrast to some other marker genes, such as luxCDABE, luc has a rather low impact on fitness.

Vertical transport of surface soil inoculated cells was studied in greenhouse columns and field lysimeters, but only the latter system yielded reliable data. In the greenhouse, more than 98% of the inoculated cells were recovered in the upper 10 cm in three of four soil columns analyzed after 85 weeks of incubation.23 However, in one column, layers below 20 cm soil depth were almost homogeneously colonized with titers of 104 to 105 cfu g-1 soil. Rim effects and crevices in the soil column probably led to transport of the surface inoculated cells in that column. Additionally, oxygen diffusion from the bottom of the soil columns, which could not completely be sealed from the greenhouse atmosphere and a homogenous temperature of the soil column, may have promoted growth in such "deeper" soil layers. In the lysimeters, the temperature and the gas atmosphere were more similar to conditions in the surrounding field soil environment. With this, more realistic system, monitoring of the inoculated cells showed no migration of the GMMs into layers below 20 cm depth (threshold of detection 100 cfu g-1 soil) and flow-through rain water did not transport any detectable bioluminescent cells through the 65 cm soil profile (threshold of detection 10 cfu ml-1). Thus, it could be concluded that no risk of vertical migration of GMMs on the field site for the subsequent field plot inoculation existed. By selection of the field site, at which the groundwater table was 20 m below the soil surface, unintentional spread into ground water could be excluded.

Horizontal spread of inoculated cells could only be analyzed in the field plot experiment. The experimental field consisted of 20 plots, each a square of 9 m2 in 4 x 5 rows, with each plot separated from the other by 3-meter noninoculated strips, seeded with grass. Plots were inoculated in block randomized order with S. melilotiwild-type, L33, L1, or not inoculated. Already 12 weeks after the field release, bioluminescent cells were detected in the rhizo-sphere of alfalfa growing on noninoculated plots.26 Two weeks later, when we analyzed the titer of bioluminescent cells in bulk soil, recombinant cells were detected on the noninoculated control plots with an average titer of 2.2 x 101 cfu g-1 soil. This titer increased further throughout the 3 year monitoring period to numbers only one order of magnitude below those on the inoculated plots.25 Mixed populations of strain 2011, L33 and L1 were found on wild-type-inoculated plots. Thus, inoculation of S. meliloti 2011 did not completely inhibit the colonization by the GMMs. Sampling outside of the alfalfa seeded plots never resulted in detection of any significant amounts of bioluminescent cells.25 The horizontal spread of the GMMs was obviously restricted to the presence of alfalfa roots.

Unintentionally, a large number of different weed plants (approx. 20 different species) grew on the alfalfa-seeded plots during the first vegetation period after inoculation. Several of such weed plants were sampled concomitantly with alfalfa plants to study their rhizo-sphere colonization by bioluminescent cells. As expected, the rhizosphere of alfalfa plants provided a well suited habitat for S. meliloti and was densely colonized by bioluminescent cells (> 105 cfu g-1 root material). Some weed plants, e.g., Capsella bursa-pastoris, did not enrich for any bioluminescent rhizobial cells, but for the weed Chenopodium album, we found approx. 103 cfu g-1 root material on inoculated and 101 cfu g-1 on noninoculated control plots, 12 weeks after the field release.

A total of approx. 1,200 pure culture colonies were isolated on growth media adapted to the isolation of rhizosphere bacteria. These isolates were obtained from rhizospheres of alfalfa and C. album plants, grown on inoculated and noninoculated plots 12 weeks after the release of the GMMs. The species richness, as detected by restriction fragment length polymorphism of PCR amplified 16S rRNA genes (ARDRA) was higher in rhizospheres of alfalfa than of C. album. The diversity of isolates was characterized at the phylogenetic level using restriction fragment length polymorphisms of PCR amplified 16S rRNA genes. The number of ARDRA pattern types, which correlated with species richness and diversity (expressed as the Shannon Index), was larger in rhizospheres of alfalfa than in that of C. album. The species richness was unaffected by inoculation in rhizospheres of C. album but increased in rhizospheres of alfalfa.26 Possibly, the S. meliloti inoculation increased the nutritional status of the early developing alfalfa plants and concomitantly resulted in the release of rhizosphere exudates stimulating the growth of a larger variety of soil bacteria.

Was this article helpful?

0 0

Post a comment