Continuous-flow slide culture offers more precise physico-chemical control of the cell environment. New medium is continuously supplied, the preventing the depletion of substrate and the accumulation of metabolic wastes. Although the chemostat approach is an improvement over batch culture, there are several potential sources of error. Traditional microculture methods, using semi-solid agar, agar blocks, or static liquid media have several disadvantages in the study of microbial behavior. Observation of predator-prey interactions and microbial competition is critical in understanding the development, structure, and function of any ecosystem. A central problem in microbial ecology is the understanding of microbial growth and behavior under in situ conditions. The development of microbial ecology has been slowed somewhat by the tacit assumption that microorganisms, like elements and molecules, have fixed properties. Microorganisms form surface biofilms in bioreactors, biological contactors, trickling filters, soil, on minerals, on particulate substrates, within the rhizosphere, in the phyllosphere, within the intestines, on the skin, in the oral cavity, and elsewhere.

Dense populations of anaerobic bacteria were found sequentially layered below the thermocline in two eutrophic lakes in southwest Michigan. Phase and electron microscopy of whole cells and thin sections were used to reveal the in situ morphology of the dominant members of the community. The predominant chlorophyll-containing bacteria were identified on the basis of their morphology to be members of the genera Pelodictyon, Prosthecochloris, Clathrochloris, Chlorochromatium, Pelochromatium, Thiopedia, Thiocystis, Thiospirillum, and Chromatium. The natural morphology of these organisms is described and compared with the morphology of reported isolates, the morphology of unisolated genera was compared with previous descriptions of natural samples. Most of the organisms near the sediment-water interface and two from the upper hypolimnion have not been previously described. They have been divided into six distinct groups based on morphology; the morphological features of each group are presented. This approach, based on the morphological uniqueness of the procaryotes present, provides a satisfactory method for grouping members of the hypolimnetic community for ecological studies.

Biological proliferation is optimized at various levels of organization, including the molecule (e.g. nucleic acids, prions), the cell (e.g. prokaryotic cells, eukaryotic cells), and the community (e.g. microbial biofilms, bioaggregates). Although it was initially assumed that this occurred through the genesis of information within DNA alone, it now appears that innovative design originates at other levels of organization in addition to DNA. For example, the recombination of community structures affects the proliferation rate of genetic structures; and the recombination of genetic structures affects the proliferation rate of community structures. This feedback mechanism computes compromises between the form and function of both community and nucleic acid. A nested series of proliferating objects (e.g. genetic structure, cell structure, community structure) is thus capable of continually updating the form of each object in the series. This accounts for the calculative nature of prokaryotic cells, eukaryotic cells, biofilms, bioaggregates, microbial consortia, and most other complex adaptive systems.

. The aim of this work was to assess the adaptation of bacterial communities to environmental transitions from labile to refractory substrates. This involved testing the hypothesis that bacteria self-organize and propagate not only as individual cellular systems, but also as functional sets of interacting organisms. A biofilm community was cultivated in a flow-cell irrigated with tryptic soy broth and subjected to a cyclic series of environmental transitions, from labile to refractory substrates, followed by a period of starvation (30 days). The appearance and disappearance of specific colony morphotypes when the emigrants were plated onto tryptic soy agar was used to monitor the restructuring of the community. Confocal laser microscopy of flow cells showed that these transitions decreased the biofilm thickness and coverage. Substrate shifts also changed the architecture of the biofilm communities. Repeated inoculation of flow-cell communities with a composite inoculum increased the number and diversity of emigrants. Their biofilms were thicker and covered a wider area than those of communities that had been inoculated only at the beginning of the experiment. With repeated inoculation, the time required for the community to restructure and stabilize decreased during most transitions. This suggested that organismal recombination acted as a mechanism of adaptation, enhancing the growth of microbial communities exposed to environmental stresses. Changes in the profiles of emigrants during the adaptation of biofilm communities to environmental transitions showed the appearance and disappearance of discrete sets of organisms. This suggested that the biofilm communities responded to environmental stresses as sets of interacting organisms. Enhanced growth of biofilm communities due to repeated environmental cycling suggested that the functionality of cellular positioning accrued from one cycle to the next and was thus heritable, although it was not necessarily genetically encoded.

Microbial organization within a biofilm community can be thought of as the product of species composition and spatial positioning of individuals within the biofilm matrix. Species composition within a microbial community, also referred to as community structure,8 determines the community’s overall genetic potential for survival and reproductive success under various environmental conditions. Spatial positioning allows individuals to interact physiologically and genetically. It also allows the creation of favorable microbial microenvironments within hostile macroenvironments. When a biofilm community is subjected to an environmental perturbation (e.g., an introduction of a pollutant or antimicrobial compound), continued reproductive success may be facilitated by a process of reorganization consisting of changes in composition and spatial arrangement of individuals within the community. Thus, the structural and spatial organization of a biofilm community, and its functional significance, should be a consideration when attempting to control or enhance the activities of biofilm communities in industrial or environmental settings.

It has been shown that microbial communities contribute extensively to the attenuation, mineralization and transport of both organic and inorganic contaminants in the environment. The development of biofilms by microbial communities is often a key factor contributing to the overall efficiency of these processes (Rothemund et al., 1996). For instance, bacterial biofilms are able to accumulate metals through various mechanisms (Marques et al., 1991; Sillitoe et al., 1994). Liehr et al. (1994) showed that biofilms formed by algae could concentrate metals at levels more than four orders of magnitude higher than those in the surrounding water.

Germ theory and pure culture methods have provided invaluable information concerning the role of bacteria in diseases resulting from a single organism which bypasses a host's defenses. However, they do not provide sufficient information concerning the synergisms which allow the members of biofilm communities to proliferate more effectively as communities rather than as individuals. The mechanisms of these synergies are potential targets for antimicrobial agents as well as potential mechanisms of resistance to antimicrobial agents. Understanding community-level phenomena in oral biology requires the culture, identification, and classification of functional plaque communities as well as new methods of identifying and quantifying communal relationships. Cultured biofilm communities also provide ideal models of bacterial self-organization in which information related to adaptive strategies arises not only through the recombination of genes within genomes, but also through the recombination of organisms within communities.

Bacterial biovolumes of hypertrophic Humboldt Lake (total dissolved solids = 3.3 g liter-1; 6 m deep) and oligotrophic Redberry Lake (total dissolved solids = 20.9 g liter-1; 17 m deep), Saskatchewan, were measured concurrently with a variety of environmental variables to identify the major factors correlated with volume changes. There was no difference (P > 0.05) in mean bacterial volume between Redberry Lake (0.084 ± 0.034 μm3 SD) and Humboldt Lake (0.083 ± 0.021 μm3 SD). Statistical analyses suggested there were marked differences in the factors associated with the pronounced seasonality of bacterial cell volumes in these two lakes. Variance in bacterial volume in the epilimnion of Redberry Lake was best explained by a multivariate regression model which included ciliate abundance and chlorophyll concentration (r2 = 0.96). The model accounting for changes in hypolimnetic bacterial volume included ciliate numbers and primary production (r2 = 0.94), of the measured variables. Bacterial volume in Humboldt Lake was most highly correlated with primary production (r2 = 0.59). Bacterial production (estimated as the rate of thymidine incorporation into DNA) and growth (thymidine incorporation rate normalized to cell numbers) were not correlated to cell volume, with the exception of cocci volume in Humboldt Lake.

The effect of gravity on the deposition of wild‐type and flagellar mutants of Pseudomonas fluorescens and Vibrio parahaemolyticus was evaluated using computer image analysis. Rates of bacterial accumulation were determined for both the upper and lower surfaces of a glass flow cell with well‐defined laminar flow conditions. Wild‐type organisms deposited on upper and lower surfaces independent of gravity, whereas flagellar mutants deposited on lower surfaces at 10 to 40 times the rate of deposition to upper surfaces. The same gravitational effect was observed for marine communities, where 30% more marine bacteria were observed to deposit on lower surfaces than to upper surfaces, and also in recolonisation studies using motile and nonmotile Pseudomonas fluorescens. These results are in contrast with earlier arguments that bacterial mass, density, and sedimentation are insignificant within flowing microenvironments. These results also showed that motility due to both polar (Pseudomonas sp.) and lateral flagella ( Vibrio sp.) can be important in overcoming gravitational forces during cell deposition processes. Computer image analysis was also used to evaluate the orientation of the longest axis of the cells with the direction of flow. The longitudinal axis of most attached cells was aligned parallel to flow lamina. Alignment was observed in all strains studied, but was greatest for nonmotile P. fluorescens mutants. Gravity had no effect on the degree of alignment for either the wild‐type or mutant strains.

The effects of glucose and nitrogen depletion on the colonization of glass Petri plates byPseudomonas fluorescens were studied in batch culture. Colonization of the surfaces was initiated before colonization of the bulk phase, and biofilm formation was observed. This resulted in an apparent lag in the batch growth curve for the cell suspension. The lag phase was an artifact caused by the partitioning of cells between the bulk and solid phase of the culture and was not due to a reduction in the growth rate of unattached cells. The specific growth rate of the unattached cells (0.331 hour−1) was almost twice that determined for the total population (0.171 hour−1). Consequently the growth rate of biofilm-forming bacteria cannot be determined in batch culture unless the growth of both attached and unattached cells is monitored, and batch growth curves may contain artifacts due to the formation and dispersion of biofilms. The depletion of either glucose or nitrogen led to the active detachment of cells from the biofilm. An increase in the hydrophobicity of unattached cells was noted on depletion of carbon. This increase was the result of emigration of cells from the surface into the bulk phase.

Computer-enhanced microscopy (CEM) was used to monitor bacteria colonizing the inner surfaces of a 1×3 mm glass flow cell. Image analysis provided a rapid and reliable means of measuring microcolony count, microcolony area, and cell motility. The kinetics of motile and nonmotilePseudomonas fluorescens surface colonization were compared at flow velocities above (120μm sec−1) and below (8μm sec−1) the strain's maximum motility rate (85μm sec−1). A direct attachment assay confirmed that flagellated cells undergo initial attachment more rapidly than nonflagellated cells at high and low flow. During continuous-flow slide culture, neither the rate of growth nor the timing of recolonization (cell redistribution within surface microenvironments) were influenced by flow rate or motility. However, the amount of reattachment of recolonizing cells was both flow and motility dependent. At 8μm sec−1 flow, motility increased reattachment sixfold, whereas at 120μm sec−1 flow, motility increased reattachment fourfold. The spatial distribution of recolonizing cells was also influenced by motility. Motile cells dispersed over surfaces more uniformly (mean distance to the nearest neighbor was 47.0μm) than nonmotile cells (mean distance was 14.2μm) allowing uniform biofilm development through more effective redistribution of cells over the surface during recolonization. In addition, motile cell backgrowth (where cells colonize against laminar flow) occurred four times more rapidly than nonmotile cell backgrowth at low flow (where rate of motility exceeded flow), and twice as rapidly at high flow (where flow exceeded the rate of motility). The observed backgrowth of Mot+ cells against high flow could only have occurred as the result of motile attachment behavior. These results confirm the importance of motility as a behavioral mechanism in colonization and provides an explanation for enhanced colonization by motile cells in environments lacking concentration gradients necessary for chemotactic behavior.

Phase, darkfield, and computer-enhanced microscopy were used to observe the surface microenvironment of flow cells during bacterial colonization. Microbial behavior was consistent with the assumptions used previously to derive surface colonization kinetics and to calculate surface growth and attachment rates from cell number and distribution. Surface microcolonies consisted of closely packed cells. Each colony contained 2(n) cells, where n is the number of cell divisions following attachment. Initially, cells were freely motile while attached, performing circular looping movements within the plane of the solid-liquid interface. Subsequently, cells attached apically, maintained a fixed position on the surface, and rotated. This type of attachment was reversible and did not necessarily lead to the formation of microcolonies. Cells became irreversibly attached by progressing from apical to longitudinal attachment. Longitudinally attached cells increased in length, then divided, separated, moved apart laterally, and slid next to one another. This resulted in tight cell packing and permitted simultaneous growth and adherence. After approximately 4 generations, individual cells emigrated from developing microcolonies to recolonize the surface at new locations. Surface colonization byPseudomonas fluorescens can thus be subdivided into the following sequential colonization phases: motile attachment phase, reversible attachment phase, irreversible attachment phase, growth phase, and recolonization phase.

Phase and computer-enhanced microscopy were used to observe the surface microenvironment of continuous-flow slide cultures during microbial colonization and to document the diversity of bacterial colonization maneuvers among natural stream populations. Surface colonization involved 4 discrete types of cell movement, which were designated as packing, spreading, shedding, and rolling maneuvers. Each maneuver appeared to be associated with a specific species population within the community. The packing maneuver resulted in the formation of a monolayer of contiguous cells, while spreading maneuvers resulted in a monolayer of adjacent cells. During the shedding maneuver, cells attached perpendicular to the surface and the daughter cells were released. The rate of growth of new daughter cells gradually decreased as the attached mother cell aged. During the rolling maneuver, cells were loosely attached and continuously somersaulted across the surface as they grew and divided. Only those populations with a packing maneuver conformed fully to the assumptions of kinetics used previously to calculate growth and attachment rates from cell number and distribution. Consequently, these kinetics are not applicable to stream communities unless fluorescent antisera are used to study specific species populations within natural communities. Virtually all of the cells that attached to the surface were viable and underwent cell division. The abundance of unicells on surfaces incubated in situ was thus primarily the consequence of bacterial colonization behavior (shedding and spreading maneuvers) rather than the adhesion of dead or moribund cells.

Computer-enhanced microscopy (CEM) was used to study the growth kinetics of bacterial microcolonies attached to the wall of a continuous-flow slide culture. Image processing increased effective microscope resolution and quantitated colony growth at 10 min intervals. Three growth parameters were used to determine growth rate: the time required for cell fission, the specific rate of increase in cell number, and the specific rate of increase in cell area. Growth rate was initially constant regardless of colony size, as assumed previously in deriving colonization kinetics. However, at low substrate concentrations growth rate varied depending on laminar flow velocity. Growth was flow-dependent at a glucose concentration of 100 mg/liter and flow-independent at a concentration of 1 g/liter. This indicated that the surface microenvironment became substrate-depleted in the absence of sufficient laminar flow velocities and that glucose rather than oxygen was rate limiting.

Online computer image enhancement and analysis were used to visualize and quantitate microbial growth in slide cultures. Images of a single field were collected, enhanced, and stored in random access memory at 15-min time intervals. The image obtained at time 0 was subtracted from each of the subsequent images. This produced a series of difference images showing microbial growth as a black object on a grey background and eliminating debris. Data obtained using difference imaging were compared to data obtained by computer analysis of the original images. The initial rate of growth (2.8 μm3∙field−1∙h−1) was determined within an error of 2% and continuously increased during the first 2 h of incubation. The images of seven soil fractions and of stationary phase cells were superimposed on both the initial and final images before subtraction to evaluate difference imaging as an aid in visually discriminating growing microorganisms from simulated artifacts. The coarse clay fraction (0.2–2.0 μm) was the most difficult to visually distinguish from microbial cells, but was easily eliminated in difference images. The potential for application of difference imaging in the direct measurement of in situ metabolic activity is also discussed.