Variability in root traits is critical in attempts to genetically improve crop performance and respond to changes in the production environment. Studying root phenomics is increasingly becoming a crop-breeding strategy. However, most of the reported studies on spinach root phenotyping mostly focused on the traits identified at one stage, but did not note how this would happen at other stages22,24. In the present study, a non-destructive RootViz FS system that could persistently track plant root growth under controlled conditions was used to quantify root system characteristics among 40 spinach accessions.
In this system, plants were grown in a low density foam medium in a controlled environment. The homogeneity of the medium and the controlled environment can help to minimise environmental error and maximise genetic differences between accessions. Simple traits such as total root length, tap root length and root width have been quickly and successfully obtained from high resolution images. In theory, the system can capture more root traits, such as root angle, but due to the particularly thin lateral roots in spinach, this cannot be achieved accurately.
The 40 spinach materials selected in this study were derived from more than 200 spinach core germplasms with good representativeness and high genetic diversity. The coefficient of variation (CV) of shoot traits varied greatly, ranging from 0.15 (PB_1) to 0.91 (OV_3), suggesting a large variation in shoot traits among spinach genotypes. Compared to shoot traits, larger CV values between accessions were observed for most of the root traits at all stages examined. It is well known that root architecture traits like RL, RA, and RTD are crucial for the plant's response to water and nutrient acquisition3. Their larger variations among genotypes have been reported in other plants, such as wheat11, sesame36, and grapevine37. All these indicated great intrinsic plasticity in spinach roots, which could allow the root phenotype to respond rapidly to environmental conditions at a given place and time38.
Among the measured root traits, the largest CVs were found in the RGR_RL, highlighting the importance of temporal root growth patterns in the analysis of phenotypic variability and genetic diversity in spinach. The dynamic patterns of root growth were also significantly reflected by the decreasing correlations with increasing sampling interval (20 and 43 DAT), possibly due to the rapid root growth during the later stages (30 and 43 DAT), which were reflected by the relatively higher values, folds, and CVs of each trait than those among the early stages. The resilience of the root system during spinach growth demonstrated the strong genotype-by-time interactions in spinach roots, which may support spinach performance across a wide range of growth conditions.
Nevertheless, considerable correlations of the same traits were found among different imaging stages, such as RL, RA, and RTD. This implied that these root-related traits have developmental relevance. Similar results were obtained by Wang and coworkers, who also reported the significant correlation between the traits studied among six vegetative stages in Brassica napus L, and suggested that root development is controlled by common genetic factors with prolonged effects9. The strong and positive correlations among RL, RA, and RTD were also in agreement with previous studies reported in wheat39, watermelon13, suggesting that these phenotypes can be used simultaneously for spinach root architecture trait improvement. Furthermore, most of the tested traits showed high stable heritability under the RootViz FS system, indicating that these traits can be selected for breeding prediction.
The strong positive correlation between root and shoot traits has been widely reported in previous studies, and certain root traits likely to improve nutrient acquisition, yield, or abiotic stress tolerance have been suggested for breeding programs as mentioned before1,40,41. Interestingly, in this study, we found that the correlations between shoot and root traits in spinach depended on their developmental stages. At the early stage, only TRL was weakly correlated with shoot traits (PH, LW, OA, and OV), suggesting that the tap root length may be more representative of the overall morphology of the early root system than the whole root size. However, root biomass-related traits, such as RFW, RL, RA, and RTD, strongly correlated with shoot biomass related traits at the later stage, indicating the more essential role of root size at the later stage than root depth in root uptake of water and nutrients for plant aboveground development.
The distinct dynamics of tap roots and lateral roots in spinach may be a mechanism for the plant's own growth or adaptation to environmental cues42. During the seedling stage, resources were allocated to primary root growth to provide sufficient space for the initiation and establishment of the first-order lateral root9. Root depth as an important root trait in plant water and nutrient uptake has been widely reported for improved biomass production43. Therefore, tap root length may play a similar role as root depth in plant establishment here. As the plants continued to grow, the distribution of resources tended towards lateral root development, and the lateral roots were the main contributors to the root biomass in spinach. This allocation between lateral and primary root growth at different stages maximized the root uptake area, thereby improving biomass accumulation in spinach. This finding was partly supported by the performance of 40 accessions, which were divided into four groups. Among them, the G2 accessions had early deeper tap roots and stronger lateral roots at later stage, as well as higher root vigor (RGR_TRL, and RGR_RL) at both early and later stages, so it is not surprising that they had the greatest biomass of root and shoot than other groups, demonstrating the importance of changes in root distribution. In contrast to G2, the accession in G1, with the lowest trait values related to tap root and lateral root, showed a lower biomass of both root and shoot. This is consistent with the previous findings that the deep and vigorous root system is ideal for achieving high yields in most field crops16.
For G3, though the accession in G3 had higher RL, RA, and RTD at 43 DAT, their shoot biomass was lower than those in G2. This is most probably caused by their significantly lower TRL at the early stage, suggesting that early weak primary roots could affect the initiation and distribution of the lateral roots, thus affecting the nutrient absorption function of the root system and the shoot biomass accumulation at the later stage. Different with G3, the accessions in G4 had relatively well-developed TRL and RL at the early stage, but their lateral root related traits at the later stage were lower than in G2, resulting in a relative poor performance of the shoot and root compared with G2. In spite of this, G4 had a relatively better performance at the early stage than other groups (except G2), which highlighted the crucial role of taproot development at seedling stage. The differentiated root performance between the G3 and G4 groups at the long-term growth stage indicated significant developmental stochasticity in plant roots. All these revealed large intrinsic temporal phenotypic variations in spinach roots, and their strong link to plant establishment should be noted. Based on these results, we proposed that early deep tap roots and later strong lateral roots can be selected as candidate traits for breeding spinach cultivars with improved resource acquisition. So, if crop yield is the main consideration for the spinach breeding target, the G2 root type may be a promising idotype for breeding, while the G4 root type with early below-ground vigor may be advantageous for quick harvesting of spinach. As for two other types of root systems, their detailed function under diverse environmental and nutritional stresses deserves to be investigated in the future.
We also found that the spinach root systems were somewhat related to their leaf shapes. The spinach with halberd-shaped leaves tended to have larger root systems than those with nearly orbicular-shaped leaves, and the accessions with strong root systems in G2 all had halberd-shaped leaves. This seems to suggest that the halberd-leaved spinach was more evolutionarily primitive and better adapted to environmental changes, resulting in a more developed root system and more rapid growth of the above-ground parts. Further experiments are needed to verify it.