**4. Discussion**

Increasing NUE is essential to ensure the environmental and economic sustainability of oilseed rape production. Nonetheless, NUE is notoriously difficult to work with and this is due to the many peripherical processes being involved. The root system is still considered as a 'black box' and very few studies have focused on oilseed rape root response to N supply. Indeed, the below-ground organ has been poorly considered in breeding programs. Today, it is recognized as a lever to improve soil resources uptake and to reduce the negative environmental impact of mineral fertilization [35,46–48]. A diversity panel of modern winter oilseed rape cultivars was challenged with divergent nitrate supplies to identify contrasting root morphologies. Such a screening strategy in laboratory conditions may speed up for breeders, the delivery of genotypes with desired root morphological features. However, given the complexity of the agricultural environment, these data obtained in laboratory conditions may not ultimately reflect the field situation [18].

In the quest for root phenotyping methods, we employed three different procedures in controlled environment and evaluated their practical values and limitations.


Overall, the three complementary culture systems permitted to observe wide phenotypic variation among the diversity panel and to discriminate between cultivars with contrasting behaviors. For example, the cultivar CLO was poorly performing across all experimental settings and N conditions, contrary to HER consistently demonstrating superior characteristics (Figures 2, 4 and 6).

Arabidopsis is a prime model organism for studying root biology. Fundamental knowledge can be translated to *Brassica* crops, which are genetically related [56,57]. The model species may serve as a resource base for studying root morphogenesis in response to N availability and N acquisition. We observed some similar root morphological responses between the two species, despite different root growth rates. Nitrate depletion stimulated the lateral root outgrowth of Arabidopsis [41] and oilseed rape (Figures 2 and 3) seedlings during in vitro culture on agar plates. However, the number of lateral roots varied less but the length of primary root (mainly LZ4) was more influenced by nitrate depletion in crop compared to Arabidopsis.

Learning about mechanisms of lateral root growth stimulation or repression by nitrate availability will help drawing strategies to optimize root system architecture. We aimed at observing the elaboration of root traits in oilseed rape cultivars and how this relates to biomass production and N capture. A premise is that root biomass production and morphological traits could be positive indicators of above-ground biomass production (and presumably yield). A rationale shared by several authors is that a branched root system that explores an important soil volume would limit nitrate leaching [18,58–61]. This study shows no negative correlation between root biomass production or total root length and shoot biomass production during in vitro and hydroponic cultures (Figures S1 and S2). Large root system deployed by BAT, JA and TRO, can support great shoot biomass production. Hence, selection for root and shoot biomass production under laboratory settings may be equally effective for improving yield.

Identifying oilseed rape cultivars with greater N capture is a pressing issue for reducing N loss in soil. This study demonstrates some potential for increased sustainability of oilseed rape production by targeting traits related to N uptake. An important variability degree for the root system size (Figure 3) and the N uptake capacity (Figure 5) was uncovered. This likely reflects no direct selection for root traits in breeding history of oilseed rape. A positive relationship was found between the total root length and the nitrate influx of HATS + LATS expressed per plant (Figure S3). This indicates that the N uptake capacity was to some extent, root-morphology dependent. Interestingly, the HATS expressed per root biomass correlated negatively with the leaf area (Figure S3). It is known that plants are modulating root N acquisition in response to shoot N demand, through long-distance mobile proteins and peptides [62,63]. That negative correlation suggests a feedback regulation of N products from the shoot on the root N uptake.
