Fig 1.
Scanning electron microscope images of the lower leaf surface of N. peltatum (A-C, summer sample; D-F, winter sample) and T. natans (G-I, summer sample). (A, D and G, 500x; B, E and H, 5,000x; C, F and I, 10,000x). a, b, c and d, bacilli; e, algae; f, cocci; and g, microbe aggregation.
Fig 2.
Total densities of bacteria and algae on the leaf, stem and root samples (A, expressed as log10 g-1 dry mass) and densities of bacteria (B, expressed as log10 cells cm-2) and algae (C, expressed as log10 cells cm-2) on the three leaf samples. Different small letters (a, b or ab) indicate significant differences among samples at p < 0.05 (Tukey’s test).
Fig 3.
Cluster analysis (based on the Bray-Curtis algorithm) and principal component analysis of all samples at genus level.
Samples attached to leaves (NSL), stems (NSS) and roots (NSR) of N. peltatum in summer, to leaves (NWL), stems (NWS) and roots (NWR) of N. peltatum in winter, and to leaves (TSL), stems (TSS) and roots (TSR) of T. natans in summer.
Fig 4.
Bacterial community structures at phylum level for all samples.
Samples attached to leaves (NSL), stems (NSS) and roots (NSR) of N. peltatum in summer, to leaves (NWL), stems (NWS) and roots (NWR) of N. peltatum in winter, and to leaves (TSL), stems (TSS) and roots (TSR) of T. natans in summer. The abundance of phyla making up less than 1% of total high-quality sequences were classified as ‘other’ in all samples.
Fig 5.
Abundance of denitrification genes in biofilm.
A and B: samples attached to leaves (NSL), stems (NSS) and roots (NSR) of N. peltatum in summer; C and D: samples attached to leaves (NWL); stems (NWS) and roots (NWR) of N. peltatum in winter; E and F: samples attached to leaves (TSL), stems (TSS) and roots (TSR) of T. natans in summer.