Rice plot, growth rate, air temperature, and ecological community dynamics.

(a) 90 cm × 90 cm rice plot and Wagner pot alignments. Four rice individuals were grown in each pot (indicated by points in the right panel of a). Heights and SPAD of the target individuals (indicated by four red points in the right panel of a) were measured everyday during the monitoring period. (b) Rice growth rate (cm/day). (c) Daily mean air temperature measured at each rice plot. Upper and lower dotted lines indicate daily maximum and minimum air temperature. (d) Ecological community compositions and average DNA copy numbers per ml water (reported in Ushio 2022). (e) The number of amplicon sequence variants (ASVs) from each water sample (reported in Ushio 2022). For b, c, and e, different colors indicate data from different rice plots.

Information transfer between rice growth and ecological community members.

(a) An example of the results of the unified information-theoretic causality analysis. Information transfer between air temperature and rice growth rates was quantified. Much higher information transfer was detected from air temperature to rice growth (left panel) compared with the opposite direction (right panel). (b) Strength of causal influence from ecological community members to rice growth. Colors indicate taxa assigned to ASVs. y-axis indicates ASV ID. Note that the prefix (e.g., “Fungi_”) of the IDs corresponds a major target group of the primer and does not necessarily indicate the taxonomic group assigned to the ASV. (c) eDNA dynamics of putative Globisporangium (Fungi_Taxa00402 in Table S3). (d) eDNA dynamics of Chironomus kiiensis (total DNA copy numbers of five midge ASVs). For c and d, different colors indicate data from different rice plots.

The manipulation experiment performed in 2019 and ecological community compositions before and after the manipulation.

(a) Setting of the manipulation experiment in 2019. The inset shows three individuals (red and green points) in each Wagner pot. Heights and SPAD of the red individuals were measured. Total eDNA copy numbers of (b) putative Globisporangium spp. and (c) midge (Chironomus kiiensis) in the rice plots. (d) Overall community compositions after the manipulation. Gray, red, and blue indicate CT (control), GN (Globisporangium nunn added), and CK (Chironomus kiiensis removed) treatments, respectively. Each ellipse indicates the overall distribution of each treatment data.

Ecological community compositions before and after the manipulation experiment in 2019.

(a) Growth rates and (b) cumulative growth of the rice individuals in the three treatment (CT = control; GN = Globisporangium nunn added; CK = Chironomus kiiensis removed) before and after the manipulation.

Differential expression genes analysis.

(a) Globisporangium nunn-added and (b) Chironomus kiiensis-removed treatment. Red and blue points indicate significant up- and down-regulated genes, respectively. Upper and lower dashed lines indicate log2(1.5) and − log2(1.5), respectively.

Examples of differentially expressed genes after the manipulation experiment.

Results of (a) Os12g0504050, (b) Os11g0184900, (c) Os01g0678500, (d) Os01g0642200, (e) Os08g0162800, and (f) Os03g0285700 are presented. y-axis represents DESeq2-normalized read counts. Gray, red, and blue indicate CT (control), GN (Globisporangium nunn added), and CK (Chironomus kiiensis removed) treatments, respectively. The gene expressions of the GN treatment in all six genes are statistically clearly different from those of the other two treatments (P < 0.0001) except for GN v.s. CK in b (P = 0.0087) and GN v.s. CK e (P = 0.00014).