*3.2. Characterization of Selected Cultivars in Hydroponics*

The vertical plate culture system delivered a rapid and extensive examination of 2D root morphology at seed germination. Next, we examined how these early observations were relating with later developmental stages. Based on the total lateral root length observed in vitro (see above), 12 contrasting cultivars were retained for further physiological analysis upon hydroponic culture (Figure 1b). Plants were fed with a nutrient solution containing 0.2 mM (N−) or 2 mM (N+) nitrate. On average for the core set cultivated in hydroponics, R (+7%) and R:S (+26%) were superior, whereas S (−24%) and R + S (−20%) were inferior at N− compared to N+ (Table S3). The leaf area (LA) (−117%) and the specific leaf area (SLA) (−66%) decreased, while TRL (+18%) and SRL (+10%) increased in the same conditions. The nitrogen concentration in root tissues (NR) (−36%) and in shoot tissues (NS) (−154%) decreased, while the carbon concentration in root tissues (CR) (+1%) and in shoot tissues (CS) (+3%) increased. The nitrogen utilization index (NUI), used as a

surrogate for the nitrogen use efficiency at a vegetative stage, increased by 56%. They were large biometric variations between cultivars (Figure 3d,e). For instance, the percentage differences between the two most extreme cultivars were in the range of 34% (AVI and REC vs. LIM) and 65% (AVI vs. HER) for LA, and of 52% (CAR vs. BAT) and 82% (EXQ vs. BAT) for TRL, respectively, at N− and N+. The responsiveness of cultivars to N depletion was also assessed (Figure 3f). A large variation in phenotypic plasticity was observed, with cultivars poorly or greatly responsive to N treatment for LA (e.g., SAV and TRO vs. RC) and TRL (e.g., CAR vs. BAT). The cultivar effect for all traits, the N treatment effect for all traits except R and NS, the interaction (cultivar × N) for all traits except R, S, SLA and TRL were significant (*p* < 0.05) (Table S3).

The two first PCs explained together three quarters of the total phenotypic variation. The PC1 (48.6%) was influenced the most by root morphological traits, while PC2 (27.1%) by the shoot biomass production (Figure 4c). The distribution of the cultivars across PC1 and PC2 permitted to clearly distinguish between the two N treatments (Figure 4d). Cultivars with low PC1 scores exhibited large leaves with elevated N concentration in tissues, while those with elevated PC2 scores produced important root biomass. We noted that the data collected from seedlings grown on vertical plates (Figure 4a,b) and those from plants grown in hydroponics (Figure 4c,d) showed the same pattern of response to the nitrate supply. The notable effect of N starvation was mainly reflected on R:S biomass ratio, TRL and SRL.

Spearman correlation coefficients were calculated between the traits measured in hydroponics (Figure S2). The strongest correlation was found between R and S biomasses (*-*<sup>2</sup> = 0.63 at N−, *-*<sup>2</sup> = 0.95 at N+). The SLA correlated negatively with CS (*-*<sup>2</sup> <sup>=</sup> −0.59 at N−, *-*<sup>2</sup> <sup>=</sup>−0.87 at N+) and NUI (*-*<sup>2</sup> <sup>=</sup> −0.79 at N−, *-*<sup>2</sup> <sup>=</sup> −0.76 at N+), and positively with NS (*-*<sup>2</sup> = 0.72 at N−, *-*<sup>2</sup> = 0.77 at N+). At N+, the LA correlated positively with R (*-*2= 0.88) and S (*-*<sup>2</sup> = 0.77). At N−, CR correlated negatively with CS (*-*2= −0.66), and SRL positively with NR (*-*<sup>2</sup> = 0.61). All reported correlations are significant (*p* < 0.001).

A nitrate influx assay with 15N tracer was performed to assess the relationship between uptake capacity of roots and plant morphological traits. The activities of the nitrate High- (HATS) and Low-(LATS) Affinity Transport Systems were measured in 10 cultivars grown in hydroponics at N+ (Figure 5).

**Figure 5.** Variability of HATS and combined HATS and LATS activities between oilseed rape cultivars. (**a**) Nitrate influx expressed per hour and per root biomass. (**b**) Nitrate influx expressed per hour and per plant. Black bars: HATS activity, white bars: HATS + LATS activity. N = 6–12 plants ± std. Bars with different letters are significantly different (*p* < 0.05) according to Tukey's HSD test.

The HATS expressed per root biomass were four-fold less important than the HATS and LATS combined values. Variations between cultivars were in the range of 200% for HATS and 25% for HATS + LATS. The AVI and CLO cultivars had significant (*p* < 0.05) greater HATS values expressed per root biomass (Figure 5a). The AVI and BAT showed, respectively, the lowest and greatest HATS + LATS values expressed per plant (Figure 5b). The HATS expressed per root biomass was negatively correlated with LA (*-*<sup>2</sup> <sup>=</sup> −0.74; *p* < 0.01), while the HATS + LATS expressed per plant was positively correlated with TRL (*-*<sup>2</sup> = 0.59; *p* < 0.05) (Figure S3).

### *3.3. In-Soil Root Phenotyping*

The X-ray computed tomography was used to examine the root system architecture of four cultivars (BAT, CLO, HER and SAV) in pipes filled with soil (Figure 1c). The profile of lateral roots along the taproot was assessed after watering with a solution without (N−) or with 50 mM KNO3 (N+) (Figure 6). The CLO cultivar was showing poor soil exploration along with poor shoot development (data not shown), and it was weakly responsive to the N treatment. It may indicate that the culture conditions were particularly unfavorable for that cultivar. On the contrary, the lateral root deployment of BAT, HER and SAV was more profuse and concentrated close to the soil surface. The root system of the BAT cultivar was repressed in the upper soil portion, while the one of SAV was clearly stimulated by N− condition. The HER genotype showed subtle profile change in response to N treatment. There was no obvious difference close to the surface but less lateral root branching down to the pipe, during N− in that cultivar.

**Figure 6.** Characterization of the root system architecture in soil-filled column. (**a**) The violin plots show the distribution of lateral root density, as a function of the column depth. The estimates are based on five replicates of four oilseed rape cultivars cultivated with low (N−) or high (N+) nitrate conditions; (**b**) Reconstructed images of two contrasting cultivars at N−. Scale bar = 5 cm.
