Bacteriophages (phages) play a crucial role in shaping bacterial communities. Documenting the relationship between phage and bacterial diversity in natural systems is fundamental to understand eco-evolutionary dynamics that shape community composition, such as host specificity, emergence of phage resistance and phage-driven microbial diversification. However, our current understanding of this relationship is still limited, particularly in animal-associated microbiomes. Here, we analyze paired bacterial and viral metagenomics data from the gut microbiota of 49 individual honeybees and reconstruct the phage-bacteria interaction network by leveraging CRISPR spacer matches and genome homology. The resulting interaction network displays a highly modular structure with nested phage-bacteria interactions within each module. Viral and bacterial alpha and beta diversity are correlated, particularly at the bacterial strain level and when considering the interaction network. Overall, our results suggest that the most relevant approach to study phage-bacteria diversity patterns should rely on strain-level resolution and the explicit use of the interaction network. This may explain why previous studies have obtained mixed results when testing for phage-bacteria diversity correlations. Finally, we call for further studies building up on these correlation patterns to probe the underlying mechanisms by considering both bottom-up and top-down regulatory mechanisms in microbiome assembly.
