Fig. 1: Ridges artificially induce mild drought and reduce soybean yield.

a Photograph of the experimental field, including the VWC data collection station, watering tubes, and probes (yellow arrows). b Twelve-day-old soybean plants on flat and ridged plots (referred to as “flats” and “ridges,” respectively). F and R denote flats and ridges, respectively. c Daily rainfall (green bars) and time course of soil VWC (n = 4 independent replicates) in the flats (blue lines and points) and ridges (pink lines and points) over the study period in the 2016 field. Relationships between soil VWC and water potential were 50.7% (−0.0039 MPa, pF 1.6), 44.9% (−0.0098 MPa, pF 2.0), 41.2% (−0.031 MPa, pF 2.5), 38.6% (−0.098 MPa, pF 3.0), and 26.3% (−0.61 MPa, pF 3.8). d Aboveground biomass (dry weight, DW) per plant grown on flats and ridges (n = 12 biologically independent replicates). e Thermogram and the corresponding digital image of 9-week-old soybean plants grown on flats and ridges in 2016. f Total seed weight per plant grown on flats and ridges (n = 20 biologically independent replicates). g Seeds from individual representative plants grown on flats and ridges. Bars, 10 cm. h Nutrient contents (N, P, and K) of soil in flats and ridges before (July) and after (November) soybean cultivation in 2016 (n = 4 independent replicates). Error bars in c, d, f, and h denote SD. *P < 0.05, two-tailed paired samples t-test (d, f). ns, no significant difference, one-way analysis of variance (ANOVA) with Tukey’s test in ridged versus flat plots for each season (h).