Fig. 5
Functional analysis grouped the microbial communities by live rhodolith, dead rhodolith, or seawater. Principal component analysis (PCA) depicting the correlations between predictor variables (SEED database level 1 subsystems - red arrows) and the metabolic similarity in rhodolith (live and dead) and seawater metagenomes for all treatments (28 metagenomes) during the experiment. The length of the arrow represents how well the parameter explains the distribution of the data. PCA grouping live rhodolith microbiome under ambient air (red circle; T1 – Tank 1) and high CO2 (dark red circle; T3 – Tank 3); dead rhodolith biofilm under ambient air (beige circle; T4 – Tank 4) and high CO2 (brown circle; T2 – Tank 2); and seawater metagenomes under four combined treatments: live algae, ambient air (light blue circle; seawater T1 - Tank 1), live algae, high CO2 (blue circle; seawater T3 - Tank 3), dead rhodolith, ambient air (gray circle; seawater T4 - Tank 4), dead rhodolith, high CO2 (dark gray; seawater T2 - Tank 2) by functional contribution in first level of hierarchy (level 1)