Biofilms, complex communities of microorganisms, are a significant concern across various industries, impacting food production, water supply, and healthcare. Their economic impact is substantial, reaching billions of USD annually. The majority of biofilms in natural and industrial settings are composed of multiple species with diverse interactions. These interactions, ranging from cooperation to competition (commensalism, exploitation, amensalism), determine community functionality and stability. In the dairy industry, multispecies biofilms can harbor pathogens and spoilage organisms, compromising food safety and quality. Bacteria surviving cleaning and disinfection (C&D) procedures often interact in ways that influence their persistence and resistance to disinfectants. Despite the importance of these interactions, our understanding remains limited due to their complexity. While pairwise interaction studies have provided insights into simplified communities, they often overlook higher-order interactions where the interaction between two species is modified by others. Keystone species, even at low abundance, significantly influence the ecology and function of other species within the community. Therefore, understanding both pairwise and higher-order interactions is essential for predicting community stability, functionality, and evolutionary dynamics. Previous work characterized 140 reproducible four-species biofilm communities from a dairy pasteuriser, identifying synergy in biofilm formation in eleven combinations. Five of these shared three species (*S. rhizophila*, *B. licheniformis*, and *M. lacticum*). The present study uses a bottom-up approach to disentangle interspecies interactions and growth dynamics in a four-species community (*S. rhizophila*, *B. licheniformis*, *M. lacticum*, and *C. indicus*) exhibiting a 3.13-fold increase in biofilm mass compared to monocultures, examining pairwise and higher-order interactions to understand the role of individual species and the significance of keystone species in community stability and coexistence.

The literature extensively highlights the importance of microbial interactions in biofilm formation and function. Studies have shown that positive interactions, such as cooperation and commensalism, are common among bacteria and contribute to community stability. Conversely, negative interactions, including competition and amensalism, can lead to community instability and reduced overall productivity. The concept of keystone species has emerged as crucial in understanding microbial ecology, as certain species can exert disproportionately large effects on community structure and function, even if they are present at low abundance. Previous research on bacterial interactions in biofilms, particularly in food processing environments, has focused primarily on pairwise interactions. However, recent studies emphasize the need to consider higher-order interactions, which involve the influence of one or more species on the interaction between two others. The importance of understanding these complex interactions for predicting community dynamics and manipulating microbial communities for various applications has been strongly advocated.

This study employed a combination of experimental techniques to characterize interspecies interactions within a four-species biofilm community. Four bacterial strains (*S. rhizophila*, *B. licheniformis*, *M. lacticum*, and *C. indicus*) isolated from a dairy pasteuriser after C&D were used. Biofilm formation was assessed using both polystyrene microtiter plates and stainless steel (SS) coupons, simulating the original environment. Biofilm mass was quantified using crystal violet staining and absorbance measurements. To investigate the role of secreted metabolites, cell-free supernatants (CFS) of each species were used to replace viable cells in various combinations. The effects of CFS on biofilm formation were compared to controls using sterile water to account for dilution effects. Bacterial growth dynamics were analyzed by quantifying cell counts on species-specific media plates at various time points (0, 4, 8, 12, 16, 20, and 24 hours). Biofilms were grown on SS coupons in both brain-heart-infusion (BHI) broth and cow's skim milk (SM) to assess the influence of different growth media. Bacterial spatial organization was observed using scanning electron microscopy (SEM). Statistical analyses (one-way ANOVA, Duncan's Multiple Range Test, paired t-tests) were performed using SPSS and GraphPad Prism to determine statistical significance.

The study revealed a strong synergy in biofilm mass production by the four-species community (3.13-fold increase compared to the sum of monocultures). Analysis of pairwise interactions revealed commensalism, amensalism, and exploitation among species. *M. lacticum* emerged as a keystone species, positively influencing the growth of all other species. Higher-order interactions were also observed, demonstrating that the addition of a third species could modify the pairwise interactions between other species. For instance, the combination of *S. rhizophila* and *B. licheniformis* showed neutral interactions in terms of cell numbers but a significant increase in biofilm mass, suggesting a cooperative effect on matrix production. Replacing viable cells with their CFS revealed that some species' metabolites could stimulate biofilm formation in other species. The growth dynamics of the four-species community in BHI and SM revealed differences in the relative abundance of each species over time. In BHI, *M. lacticum* dominated initially but declined after 12 hours while *S. rhizophila* and *B. licheniformis* increased. In SM, *M. lacticum* remained stable for a longer period, potentially affecting the competitive balance within the community. pH measurements indicated that *M. lacticum* decreased the pH, while *S. rhizophila* and *B. licheniformis* increased it. These pH alterations likely contributed to species-specific growth patterns. A schematic representation of the interaction network revealed that *M. lacticum*'s growth-promoting effect mediated several pairwise interactions, highlighting the complex interplay within the community.

The findings demonstrate the importance of both pairwise and higher-order interactions in shaping the structure and function of this multispecies biofilm community. The keystone role of *M. lacticum* highlights the disproportionate impact a single species can have on community dynamics. The observation that the outcomes of pairwise interactions could not fully predict the behavior of the four-species community emphasizes the significance of higher-order effects. The use of different media (BHI and SM) underscores the influence of environmental factors on bacterial interactions and community stability. The discrepancies observed in growth patterns and species abundance between BHI and SM suggest that nutrient availability is a crucial determinant of the competitive balance and the role of keystone species. These results have implications for understanding microbial community dynamics in natural and industrial settings, particularly for food processing environments. Targeting keystone species like *M. lacticum* may offer novel strategies for controlling and manipulating biofilms.

This study provides novel insights into the complex interplay of dynamic social interactions and keystone species in shaping the diversity and stability of mixed-species biofilms. The keystone role of *M. lacticum* in promoting biofilm synergy, the importance of higher-order interactions beyond pairwise effects, and the influence of environmental factors are crucial considerations for understanding biofilm dynamics. Future research could focus on gene expression profiling to understand the molecular mechanisms underlying these interactions and explore the potential for manipulating these interactions to control biofilms in industrial settings.

The study primarily focused on a specific four-species biofilm model. The generalizability of the findings to other biofilm communities requires further investigation. While the use of CFS provided insights into the role of secreted metabolites, it might not fully capture the complexity of interactions in the structured biofilm environment. The dilution effect in CFS experiments could also have affected the interpretation of results. Further research is needed to explore the impact of fluctuating environmental conditions and long-term community stability.