Fig. 3

Rhythmicity of bacterial and fungal OTUs under diurnal light–dark cycles. A–D Venn diagrams comparing OTUs identified as rhythmic based on comparison of dawn and dusk samples in the rhizosphere of different plant genotypes or in soil, based on analyses of total or active microbial communities. Lists of rhythmic OTUs detected in this experiment are provided as Additional File 1. E Experimental design for time-course sampling. F, G Expression of the circadian clock genes CCA1 and LUX in root samples from the time course. White and shaded areas indicate days and nights, respectively. BHQ values indicate Benjamini–Hochberg corrected p values for rhythmicity. H Raster plots displaying temporal changes in the relative abundance of individual bacterial and fungal OTUs in soil and rhizosphere samples. Each rectangle represents the relative abundance of one OTU at a particular time point, from one sample derived from 3 to 4 plants. A moving 2-point average was applied to reduce noise in the data. Each row of panels shows OTUs that were detected as rhythmic by MetaCycle in the rhizosphere of wild-type, lhy-11 and lhy-ox plants or in bulk soil, based on period values between 18 and 30 h, and BHQ < 0.01. Each column shows temporal patterns for these same OTUs in the other samples. OTUs were sorted according to period length and phase in the sample in which they were detected as rhythmic. White and black rectangles below the chart indicate days and nights respectively. Lists of rhythmic OTUs detected in this experiment are provided as Additional File 3