In pure culture biofilms, 80% of the biofilm weight is occupied by exopoly­saccharides (EPS) (Nelson et al., 1996). In continuous-flow systems, shear forces at the liquid/biofilm surface cause cell detachment and loss of EPS. Viable cells in the biofilm matrix must continuously replace the displaced biofilm materials thereby drawing on the cell’s energy resources. Additionally, cell attachment plays an important role in the cell division cycle inside the biofilm. For example, Meadows (1971) observed that Pseudomonas fluore- scens and Aeromonas liquifaciens cells undergo cell division only during their most stable attachment phase (when lying longitudinal to the solid surface). In more dramatic cases such as the prosthecate bacterium Caulobacter crescentus, cell division occurs only during the attachment phase of the cells life cycle (Neidhardt et al., 1990). In Caulobacter, the prosthecate (stalked) form undergoes cell division giving rise to a swarm cell equipped with a flagellum while the other daughter cell remains attached to the surface with a single prostheca (stalk). The life cycle is completed by differentiation of swarm cells to prosthecates, followed by attachment and cell division.

Scientists and mathematicians have found biofilms too complex to analyze and model especially when working with mixed-culture communities. Biofilm systems in laboratory studies are often oversimplified by using systems with defined chemical and microbial species composition. Attempts to use data from natural systems to develop the models have been unsuc­cessful because of the variability and complexity of the natural aquatic environments. Recently, biofilm studies using microsensors have unravelled the existence of closed-cycles, such as the cycle of sulfate oxidation coupled with sulfate reduction, within biofilms or sediments (Kuhl and Jprgensen, 1992). Models based only on bulk liquid and effluent characteristics often underestimate the overall performance of biofilm systems. The internal metabolic cycles in the biofilms, though hidden from the external observer, determine the final structure of the microbial community thereby affecting the long-term performance of the biofilm (Santegoeds et al., 1998).

Furthermore, bacterial attachment and conglomeration play an impor­tant role in the survival of cells under hostile conditions. For example, attached cells may feed on adsorbed substrates on the surfaces of the immo­bile phase under starvation concentration levels (Zobell, 1943). In toxic environments, biofilm communities may be exposed to lower levels of toxic­ity either due to the masking effect of less susceptible species or due to heterogeneities within the biofilm microenvironment (Chen and Stewart, 1996; Nichols, 1989). Several researchers have proposed mass transport resistance as the main mechanism limiting penetration of toxic substances into biofilms (Xu et al., 1998; Chen and Stewart, 1996; Hodges and Gordon, 1991; Hoyle et al., 1992).

In engineering attached growth bioreactors, the biofilm is allowed to grow thick. This results in a complex dynamic at the surface with a high rate of cell shading and sloughing of the biomass. Inside a thick biofilm, tunnel­ling and honeycombing occurs which makes the prediction of the actual surface area for interfacial diffusion impossible. For this and the above reasons, biofilm models are regarded as mere approximations of the pos­sible operating conditions.