Distribution and population dynamics of Rhizobium sp. introduced into soil

anglais. In this thesis the population dynamics of bacteria introduced into soil was studied. In the introduction, the existence of microhabitats favourable for the survival of indigenous bacteria is discussed. Knowledge about the distribution of introduced bacteria over such microhabitats, however, is scarse. Nevertheless, it was hypothesized that upon introduction, bacteria reach other microsites in soil than bacteria which are already present for some time, thereby influencing the survival of introduced organisms. Methods to study the distribution of introduced bacteria in soil, as well as the effect of their distribution on the population dynamics, were assessed. A model organism, Rhizobium leguminosarum biovar trifolii and two different soils, a loamy sand and a silt loam, were used for this purpose.Two methods for the enumeration of bacteria introduced into soil were compared (Chapter 2). Although immunofluorescence was a very promissing method at the moment we started our work, selective plating proved to be more suitable for the enumeration of low numbers of introduced bacteria, since it had a lower detection limit than immunufluorescence. In addition, selective plating did not depend on flocculation processes which were shown to influence significantly the results obtained with the immunofluorescence technique. Moreover, only cells able to divide were counted. However, with the immunofluorescence technique we were able to determine cell lengths and we detected that the length of cells which were grown in a rich medium decreased after their introduction into soil.To study the micro-distribution of bacteria in soil, different fluorochromes were tested on their ability to stain bacteria in thin sections of undisturbed soil samples. Calcofluor white M2R in combination with acridine orange was successfully applied for the detection of bacteria in thin soil sections (Chapter 3). However, specific staining of the introduced rhizobia with conjugated antiserum was not successful. Therefore, an alternative method for the assessment of the distribution of introduced bacteria, a soil washing procedure, was used in Chapters 4, 5 and 6. With this method, free occuring bacteria and bacteria associated with soil particles or aggregates>50 μm were distinguished.The bacterial distribution through soil could be manipulated by inoculating soils at different initial moisture contents. At a lower initial moisture content, only the narrowest pores are filled with water, so that inoculated rhizobial cells will reach narrower pores when they are transported passively by the waterflow. At a higher initial moisture content, water already present in the narrower pores prevent the introduced cells from entering these pores. With such an inoculation procedure, rhizobial cells were found to be associated to a larger extent with soil particles when soils were inoculated at lower initial moisture contents. In natural soils, this resulted in an improved survival of rhizobia during at least 100 days after inoculation (Chapter 4). Moreover, the number of particleassociated cells decreased less than the number of free occuring cells in natural soil. It was concluded that rhizobial cells associated with soil particles or aggregates>50 μm occupied a more favourable microhabitat than free occuring cells. In sterilized soil, numbers of both particleassociated and free occuring cells increased and the initial differences in distribution did not result in different final population levels (Chapter 5). Therefore, it was concluded that the microhabitats in natural soil rendered protection to biotic rather than to abiotic factors.The influence of competitors and predators on the distribution and population dynamics of rhizobium was studied by the addition of specific groups of organisms to sterilized soils (Chapter 5). Previous to inoculation with rhizobia, sterilized soils were recolonized with several bacterial isolates which were obtained from these soils and part of the soil portions were inoculated with a flagellate precultered on rhizobial cells. In the presence of flagellates, which predate on bacteria, a higher percentage of particle-associated rhizobial cells was present than in the absence of flagellates. In recolonized soils, i.e. in the presence of competitors, the percentages of particle -associated rhizobial cells were lower than in soils that were not recolonized previous to inoculation. Thus, the presence of competitors made it more difficult for rhizobial cells to colonize the microsites where they can be associated with soil particles or aggregates. The total number of rhizobial cells was influenced only little (silt loam) or not at all (loamy sand) by the competitors or by the addition of flagellates alone. However, when both competitors and predators were present, numbers of rhizobial cells decreased drastically. This synergetic effect was explained by hypothesizing that after the predation of accessible bacterial cells by the flagellates, the regrowth of rhizobial cells will be limited by the presence of competitive microorganisms in many of the favourable microhabitats.The association of rhizobial cells with soil particles may be the result of enclosure in pores or attachment to soil surfaces of rhizobia. The role of attachment was studied with a R. leguminosarum strain and three Tn5 mutants which were altered in their cell surface properties (Chapter 6). Although the importance of association with soil particles or aggregates was affirmed, the results gave no evidence that attachment to soil particle surfaces was an important factor for the survival of introduced cells.The final population level of introduced rhizobia was studied in more detail by inoculating sterilized and natural soils with different inoculum levels (Chapter 7). In sterilized soils, populations reached, independent of the inoculum density, a final level which was suggested to represent the carrying capacity of the soils in terms of available habitable pore space, moisture and substrate for survival of the bacteria. In natural soil, however, the survival levels were dependent on the inoculum density. In this case, the chances of introduced cells to reach favourable microhabitats, determined the survival level of the entire population.In all experiments final population levels in natural and in sterilized soils were higher in the silt loam than in the loamy sand (Chapter 2-8). Pore space which is suitable for bacteria to survive (=habitable pore space) or which protects bacteria from predation (=protective pore space) was estimated for both soils. The occupancy by bacteria was in all cases lower than 0.5%, so that no serious space limitation could be expected. Therefore, the larger water-filled pore volume at the water potential used (pF 2) in the silt loam as compared to the loamy sand, could not explain the differences in population sizes. In sterilized soil substrate availability was suggested to determine the final population level. In natural soil, however, the survival of rhizobial cells was suggested to be dependent on the number of introduced bacteria that reached the protective pore space (Chapter 4 and 7).In this thesis it is shown that the soil washing procedure is useful for the study of the distribution of introduced bacteria. Immediately after introduction, only few bacteria were associated with soil particles. The number of particle-associated bacteria decreased less pronounced than the number of free occuring bacteria, giving evidence that the distribution of introduced bacteria in soil is indeed an important factor influencing its survival. Moreover, the distribution could be manipulated by inoculating soil at different moisture contents. Inoculation of dryer soils, as well as the use of higher inoculum densities, resulted in higher survival levels, which could be well explained with the concept of distribution of cells over protective and non-protective pore space. The occurance of different population levels under apparently the same environmental conditions during incubation, suggests that extensive translocation in natural soil is absent.The ability to manipulate the distribution and thereby to influence the survival of introduced bacteria, is important for the application of bacteria in soil. The availability of methods for biological control of soil-borne pathogens, nitrogen fixation and degradation of xenobiotics in soil, is depending on the possibility of introduced bacteria to establish in soil. The knowledge obtained in this research project can be used to improve the survival of bacteria introduced into soil.The spatial distribution of bacteria through the soil matrix might also be a useful concept for other areas in soil(micro)biology. The occurance, for example, of genetransfer in different soil systems might be better understood when more details about bacterial distribution are available. Also the preservation of organic matter and the activity of predators will depend on the distribution of bacteria in the soil matrix.

néerlandais; flamand. In eerste instantie is in deze studie gezocht naar een geschikte methode om de ruimtelijke verdeling van in de grond geintroduceerde bacterien te onderzoeken i.v.m. hun overlevingskansen. En vervolgens is de invloed van de ruimtelijke verdeling van deze bacterien op hun overleving bestudeerd aan de hand van een model-organisme, Rhizobium leguminosarum biovar trifolii, uitgevoerd in de beide grondsoorten zwak lemig zand en kleigrond.