TABLE 2.

Variance explained by categorical predictors when applied to the entire data set and core assignments^a

Method	Categorical predictors
	Human Microbiome Project data set						Arabidopsis data set
	Bray-Curtis				Binary		Bray-Curtis		Binary
	Visit no.	Patient sex	Sequencing center	Visit no.	Patient sex	Sequencing center	Developmental stage	Genotype	Developmental stage	Genotype
Full dataset	P < 0.01, R2 = 0.004	P < 0.001, R2 = 0.005	P < 0.001, R2 = 0.080	P < 0.05, R2 = 0.003	P < 0.01, R2 = 0.004	P < 0.001, R2 = 0.058	P < 0.001, R2 = 0.032	P < 0.001, R2 = 0.052	P < 0.001, R2 = 0.014	P < 0.001, R2 = 0.036
Co-assigned core by all methods	P < 0.01, R2 = 0.005	P < 0.05, R2 = 0.004	P < 0.001, R2 = 0.120	P > 0.1, R2 = 0.002b	P > 0.05, R2 = 0.004b	P < 0.001, R2 = 0.071	P < 0.001, R2 = 0.043	P < 0.001, R2 = 0.062	P < 0.001, R2 = 0.035	P > 0.1, R2 = 0.032b
Abundance-based	P < 0.05, R2 = 0.004	P < 0.01, R2 = 0.005	P < 0.001, R2 = 0.090	P > 0.1, R2 = 0.003b	P < 0.01, R2 = 0.005	P < 0.001, R2 = 0.069	P < 0.001, R2 = 0.039	P < 0.001, R2 = 0.059	P < 0.001, R2 = 0.024	P < 0.001, R2 = 0.045
Occupancy-based	P < 0.05, R2 = 0.006	P < 0.05, R2 = 0.005	P < 0.001, R2 = 0.119	P > 0.1, R2 = 0.003b	P > 0.05, R2 = 0.004b	P < 0.001, R2 = 0.069	P < 0.001, R2 = 0.038	P < 0.001, R2 = 0.059	P < 0.001, R2 = 0.020	P < 0.001, R2 = 0.045
Abundance and occupancy-based	P < 0.01, R2 = 0.006	P < 0.05, R2 = 0.005	P < 0.001, R2 = 0.080	P > 0.1, R2 = 0.003b	P > 0.05, R2 = 0.004b	P < 0.001, R2 = 0.069	P < 0.001, R2 = 0.039	P < 0.001, R2 = 0.061	P < 0.001, R2 = 0.021	P < 0.001, R2 = 0.047
Hard cutoffs of abundance and occupancy	P < 0.05, R2 = 0.004	P < 0.01, R2 = 0.005	P < 0.001, R2 = 0.094	P > 0.5, R2 = 0.002b	P < 0.05, R2 = 0.004	P < 0.001, R2 = 0.069	P < 0.001, R2 = 0.043	P < 0.001, R2 = 0.062	P < 0.001, R2 = 0.033	P > 0.1, R2 = 0.033b

^aR² values indicate the variance explained by a categorical predictor, and thus its importance.

^bSignificance of predictor differs between core assignment and the full data set.