Belowground plant development measured with magnetic resonance imaging (MRI): exploiting the potential for non-invasive trait quantification using sugar beet as a proxy

Abstract

Both structural and functional properties of belowground plant organs are critical for the development and yield of plants but, compared to the shoot, much more difficult to observe due to soil opacity. Many processes concerning the belowground plant performance are not fully understood, in particular spatial and temporal dynamics and their interrelation with environmental factors. We used Magnetic Resonance Imaging (MRI) as a noninvasive method to evaluate which traits can be measured when a complex plant organ is monitored in-vivo while growing in the soil. We chose sugar beet (Beta vulgaris ssp. vulgaris) as a model system. The beet consists mainly of root tissues, is rather complex regarding tissue structure and responses to environmental factors, and thereby a good object to test the applicability of MRI for 3D phenotyping approaches. Over a time period of up to 3 months, traits such as beet morphology or anatomy were followed in the soil and the effect of differently sized pots on beet fresh weight calculated from MRI data was studied. There was a clear positive correlation between the pot size and the increase in fresh weight of a sugar beet over time. Since knowledge of the development of internal beet structures with several concentric cambia, vascular and parenchyma rings is still limited, we consecutively acquired 3D volumetric images on individual plants using the MRI contrast parameter T2 to map the development of rings at the tissue level. This demonstrates that MRI provides versatile protocols to non-invasively measure plant traits in the soil. It opens new avenues to investigate belowground plant performance under adverse environmental conditions such as drought, nutrient shortage, or soil compaction to seek for traits of belowground organs making plants more resilient to stress.

Keywords: Beta vulgaris ssp. vulgaris (sugar beet); cambial rings; imaging (3D); magnetic resonance imaging (MRI); non-invasive method; root development.

Figure 1

Comparison of a sugar beet plant (Beta vulgaris ssp. vulgaris cv. “Pauletta”) photographed before and after harvest and imaged with Magnetic Resonance Imaging (MRI). (A) Photograph of the plant 118 days after sowing (DAS) growing in a soil-filled container with an inner diameter (I.D.) of 117 mm and a total height of 800 mm (only the upper part shown). (B) Volume rendering of MRI data showing both the outer shape of the beet and side roots in the soil. Different parts of the beet are indicated by yellow lines: the head of the beet (h) with the onset of petioles, the transitions zone of the neck (n), and the root part (r). A groove with rootlets (^*) and some side roots (arrowheads) were visible as in (C), an optical image of the same beet taken after excavation. The MRI image was obtained with a three dimensional spin echo sequence. The image size was 128 × 64 × 256 voxels (256 = vertical direction) with a field of view of 70 × 60 × 140 mm³ resulting in a voxel resolution of 0.54 × 0.94 × 0.54 mm³. Signal loss toward the top and bottom of the beet was caused by a loss of radio frequency (RF) homogeneity toward the upper and lower end of the RF coil. Scale bars: (A) 50 mm; (B,C) 10 mm.

Figure 2

Series of MRI images showing the development of a sugar beet in the soil with largest diameter highlighted. (A–E) Volume renderings of 3D volumetric MRI datasets between 53 and 130 DAS. The sugar beet plant was grown and measured in a container with 81 mm I.D. and a height of 400 mm. Yellow planes denote the position of the largest diameter of the beet. In (A–D) the field of view (FOV) was 70 × 50 × 140 mm³ with an image size of 128 × 48 × 256 voxels and, in (E) the FOV was 70 × 70 × 140 mm² with an image size of 128 × 64 × 256 voxels. The same measurement protocol as in Figure 1B was used. Scale bar: 20 mm.

Figure 3

Development of sugar beets grown in containers with either 117 mm I.D. (800 mm in height; cf. Figure 1) or 81 mm I.D. (400 mm in height; cf. Figure 2) for which calculated fresh weights of the beets (cFW-117 and cFW-81) were derived from MRI measurements every second week. A polynomial of 2nd order was fitted to the data points (for both R² = 0.9998). The fresh weight of the beets in the large containers was taken after harvest following the respective last MRI measurement at 118 DAS (FW-117). For the small containers, the cFW-81 curve beyond 129 DAS (dotted line) was extrapolated based on the polynomial for comparison with the fresh weight taken at slightly delayed harvest on 139 DAS (FW-81). The MRI measurements used the same MRI sequences and parameters as of Figure 1B. The values and error bars are given as arithmetic mean ± SD (n = 8 plants, respectively).

Figure 4

Internal structures of different regions of a sugar beet imaged with MRI at 129 DAS and light microscopic images of cross sections taken from another plant of similar age. (A) Volume rendering of the 3D dataset of the whole beet, with virtual cuttings to visualize internal longitudinal structures. Colored planes (yellow and blue) denote the positions of the virtual cross sections of (B,D). Magnified sections of (B,D) are shown in (C,E), respectively. The cross sections represent maps of the MRI contrast parameter T₂ (transverse relaxation time) on virtual slices through the beet scaled from 0 to 35 ms; for clarity of view, T₂ times >35 ms were set to 35 ms. For identification of the structures indicated by arrows in the MRI images they were highlighted in the microscopic images (stained with Astra Blue and Safranin), with (F) showing an overview including the beet core and three rings and (G) displaying higher magnification of a typical vascular bundle of such cambial rings. Identified tissues include parenchyma (Pa), cambium (Ca and white arrowheads), xylem (X and red arrowheads), and phloem (Ph and blue arrowheads). Plants were grown in tubes with 81 mm inner diameter. Image (A) was acquired with the MRI protocol of Figure 1B but with slightly modified parameters (FOV 70 × 70 × 140 mm³ with an image size of 128 × 64 × 256 voxels). The T₂ maps of (B,D) were achieved with a multi-echo sequence (slice thickness in vertical direction 1 mm, FOV in plane 63 × 63 mm³ with an image size of 384 × 384 pixels). Scale bars: (A–E) 5 mm; (F) 1 mm; (G) 0.1 mm.

Figure 5

Development of the rings of a sugar beet at the thickest part of the root region visualized and analyzed by MRI images (T₂ maps). The virtual cross sections (A–D) show the development of the beet and the rings between 81 and 129 DAS. Magnified areas at 81 (B) and 129 DAS (D) are presented in (E) and (F), respectively. The development of the width of the innermost four rings over time is shown in (G), including 53 DAS for which the T₂-map is not shown here. Ring width was measured as the distance between the cambia. For identification of the particular tissues, abbreviations and measurement parameters see Figure 4. Scale bars: (A–D) 10 mm; (E,F) 5 mm.