Eukaryotes have a microbiome, which confers a variety of physiological processes. When this relationship is disrupted, the host can suffer altered physiological functioning, and can even mortality. Inherent microbiome variability has however challenged our view for what constitutes a healthy microbiome, therefore suggesting the microbiome interacts with the host beyond the presence of specific microbiome members. Co-occurrence patterns among microbiome members can also influence host health through emergent microbiome function; however, the relationship between co-occurrence patterns and diversity structure of the microbiome are not known. Here, we analysed the microbiome from the epidermal surface of Rhincodon typus (whale sharks), across distinct oceanic regions, including Mafia Island, Tanzania; Ningaloo, Australia; Oslob, Philippines; and La Paz and Isla Mujeres, Mexico. Microbial taxonomic and gene function compositions were described using shot-gun metagenomics. Alpha and beta diversity patterns were distinct, across five sites from each major ocean basin. The highest and lowest taxonomic alpha diversity occurred on sharks from Ningaloo (eH` = 47.9 ± SD 4.3) and Cancun (20.4 ± 1.2), while gene function diversity was greatest at Tanzania (86.5 ± 0.8) and lowest at Ningaloo (82.5 ± 0.7). Taxonomic (F df = 4,70 = 4.77; R2 = 0.21; p < 0.01) and gene function (F df = 4,70 = 28.05; R2 = 0.61; p < 0.01) beta diversity was also distinct, with location having a larger effect on gene function relative to taxonomic beta diversity. Despite distinct diversity patterns, the microbiomes of the whale sharks from each location had similar network architecture for taxonomic and gene function patterns. Therefore, the microbiomes, while having distinct diversity patterns appear to have similar microbiome complexity, indicating that microbes of the microbiome arrange their abundance patterns in a similar manor to hosts from different locations. We suggest network architecture represents a fundamental microbiome attribute that is more revealing of host-microbiome interactions than diversity patterns alone, potentially being more informative towards understanding the host-microbiome health relationship.