Rumen microbiome composition of beef cows across Córdoba, Colombia characterized by indicator analysis, microbial networks, and predictive modeling

Geographic structuring of the rumen microbiome has practical implications for region-adapted management in beef systems, yet often occurs against a large, shared community background. To investigate this, rumen microbiomes from Colombian cattle across six Córdoba subregions were profiled using 16S rRNA gene sequencing to test for subregional differences in diversity, composition, discriminant taxa, and co-occurrence structure. Analysis revealed a prominent core microbiome, with 18,032 ASVs shared across all subregions alongside subregion-specific ASVs. Alpha diversity was high but did not differ significantly among subregions. In contrast, community composition differed by subregion: Bray-Curtis and Unweighted UniFrac analyses showed significant regional effects explaining ~9-10% of variance. Multivariable ordination indicated that region was the strongest driver among measured factors and retained an independent effect. Taxonomic profiles were dominated by Bacteroidota and Firmicutes, with Sabana showing a distinct shift toward higher Firmicutes. Indicator-family analysis identified subregion-associated taxa (e.g., Pseudomonadaceae/Erwiniaceae in Bajo Sinú; Bifidobacteriaceae in San Jorge; Bacteroidales_UCG-001 in Alto Sinú), and supervised classification recovered a limited but detectable discriminatory signal. Co-occurrence networks also differed significantly in density, with denser networks in Bajo Sinú and Sinú Medio and more modular structure in Alto Sinú and Sabana. This study provides the first subregion-resolved baseline characterization of the rumen microbiome in beef cattle from Córdoba, Colombia, and demonstrates that geographic differences persist as an independent signal. By integrating indicator analysis, supervised classification, and co-occurrence topology, a compact, reproducible set of region-linked families was identified that are not only discriminant but also network-central, offering tractable candidates for functional validation and region-tailored microbiome management.