The total bacterial and fungal functional richness was recently predicted to be millions of KEGG level 3 functions but due to limited space and resource availability together with constraining environmental conditions, the local functionality only comprises a subset of the total functionality. However, widespread functions whose abundance and redundancy depends on species richness or sequencing depth could bias functional indices. Here, we used 20 randomly chosen extant metagenomes from low to high species richness each from aerial, aquatic and terrestrial ecosystems to quantify the distribution of widespread KEGG level 3 functions and their buffering effect on diversity, redundancy and richness.
A different number of widespread functions were detected in each ecosystem with 1,201 found in all three. Diversity found to be similar in the three ecosystems and redundancy that was significantly lower in aerial metagenomes showed identical patterns when all functions were used or when the widespread functions were removed. However, functional richness changed from being significantly higher in aquatic than in terrestrial and aerial metagenomes when all functions are used to no significant differences when the widespread functions are removed. The variance in diversity and richness was higher without widespread functions, making it possible to compare highly dense environments that would otherwise result in similar values.
Taken together, we describe the widespread functiome in aerial, aquatic and terrestrial metagenomes, and its buffering effect on functional indices that are driven by widespread functions which is why, moving forward, we recommend to remove widespread functions from the analysis of metagenome functionality.