Most people are familiar with the phrase ¨A picture is worth a thousand words¨, and especially in the scientific field, it is useful because researchers need to “demonstrate their hypothesis”. Consequently, a good way to do it is through an image.
Plant imaging is a growing area of biology that allows the assessment of complex plant traits such as growth, development, tolerance, resistance, architecture, physiology, ecology, among others, through tools and methods that include visual features.
In this article, you will see the multiple scale high-resolution imaging commonly used in plant sciences to assess plant traits at different biological scales like molecule to cell, organs, whole plant, and canopy.
In this article, you will see:
Plant phenotyping refers to a quantitative description of the plant’s anatomical, ontogenetical, physiological and biochemical properties (Walter et al., 2015).
But, how is plant phenotyping related to plant imaging? Nowadays, rapid developments are taking place in the field of non-destructive tissue techniques, which uses image-analysis-based phenotyping that allows for high-throughput characterization of plant traits.Modern, high-resolution imaging techniques facilitate the visualization of multi-dimensional and multi-parameter data. Imaging is used in plant phenotyping to quantify complex traits such as growth, yield, and stress conditions in both controlled environmental systems (growth chambers or in the greenhouse) or the field.
Imaging from Molecule to Cell
A range of techniques have been developed to operate at the scale from molecule to cell, between 10nm to 10μm, with different parameters being analyzed. Below, I described the image techniques and their applications at molecule and cell levels.
In visible-light imaging, cameras sensitive in the visible spectral range are used to create digital images. Cameras work as sensors sensitive to visible bands of light (400–750 nm), creating two dimensions (2D) images. The information is presented as intensity values corresponding to the number of photons (in µmol) per second and unit area on a surface and given (also known as photon fluxes) in the red (~600 nm), green (~550 nm), and blue (~450 nm) spectral bands of visible light. This configuration of red (R), green (G), and blue (B) in the cameras are known as RGB.
Currently, optical super-resolution imaging has improved the visualization of molecules and plant cells at higher levels. Some pigments have natural colors; therefore, visible-light imaging can quickly reveal the position for those pigments. Furthermore, at the cellular level, some of the phenotyping parameters accessible to study with visible-light imaging include cellular morphology and architecture and localization of cell types within a tissue. Cells within a tissue can be further differentiated using dye-based approaches.
Cameras used in fluorescent imaging can obtain information about a plant's metabolism. Here, the information about a plant’s metabolic status can be obtained by the artificial excitation of the plant photosystems and observation of the relevant responses. Fluorescence is light that is emitted during the absorption of radiation in some shorter wavelengths. Typically, the fluorescing part of the plant is the chlorophyll complex. When the chloroplasts are irradiated with blue or actinic light, part of the light is re-emitted. This re-emitted light is the fluorescence.
At the molecular level, fluorescent proteins are fused to a protein of interest transiently or stably to visualize the fusion protein's expression, organization, and dynamics in vivo using various live microscopy techniques. In addition, biosensor FPs have been used successfully in plants for reporting on a wide variety of metabolites, ions, and intracellular processes, such as glucose, GTPase, cAMP, Ca++, pH, and phosphorylation.
At the cellular level, dead plant cells can be differentiated from living cells thanks to fluorescent markers like Fluorescein Diacetate or FDA. Here, live cells appear green under the microscope, while dead cells do not express any fluorescence.
Some specific imaging techniques from molecules to cells include photoactivated localization microscopy (PALM), stochastic optical resolution microscopy (STORM) and stimulated-emission depletion microscopy (STED). These techniques are based on super-resolution, improving the deep look into the plant cells and molecules, outperforming the traditional wide-field microscopes.
Visible imaging is most often used at a biological scale from cell to organs (0.1 μm to dm) to see growth dynamics, yield and flower traits, shoot biomass, root architecture, leaf morphology, seedling vigor, Imbibition and germination rates, organ length, and seed morphology.
Furthermore, fluorescence imaging is also used at this scale to estimate photosynthesis, monitor the effects of plant pathogens and diagnose early stress responses to abiotic and biotic factors before a decline in growth. Fluorescence imaging provides a rapid screening technique to identify plants with improved or impaired metabolism and growth.
Near-infrared cameras are used to map the surface temperature of the cells, allowing the visualization of infrared radiation when detecting the temperature across the object’s surface.
Parameters such as surface temperature, stomatal conductance or water stress (induced by biotic or abiotic factors) can be measured with thermal imaging.
Laser scanning instruments with widely different ranges are used to perform plant phenotyping. A pulsed laser is used to illuminate the target (in this case the cell or the organ). Through imaging detection of the target's reflected light, more detailed information of the target can be obtained: such as shape, morphology, position, and other physical characteristics (Fan et al 2020).
Laser imaging is used to measure shoot biomass and structure, leaf angle distributions, root and stem height, and architecture to create 2D images.
Specific techniques at this level include microscopy techniques like X-ray phase contrast imaging (X-ray PCT), sheet fluorescence microscopy (LSFM), multi-angles confocal microscopy and optical projection tomography (OPT). These techniques allow for the analysis of the microstructure in whole seeds, observe the entire seedling growth cell by cell, or image a whole leaf with the possibility of cellular level resolution.
Researchers also use visible light imaging to evaluate multiple organs at the morphological and architectural level from μm to m. Fluorescence imaging can determine the leaf health status by measuring photosynthetic-associated characteristics (intense fluorescence means a good photosynthetic status) and shoot architecture. Furthermore, laser imaging may also be used to see root and shoot structure and biochemical changes of these organs.
Stereo camera systems and time-of-flight cameras can be used to create 3D images. Similar to laser imaging, 3D imaging can determine root and shoot architecture (width and height) and leaf angle distributions; however in greater detail because of the additional perspective of the 3D image.
Specific experimental setups such as the rhizotron allow studying the root and its interactions with the environment. Furthermore, the rhizotron can be combined with fluorescence imaging to measure the root growth and differentiate meristem cells (mother cells) from the rest of the organ.
Researchers can visualize and understand relationships at canopy level from meter (m) to hectometer (hm). Plant imaging spectroscopy is a technique which enables the measuring and capturing of information from the canopy in a remote way (remote sensing).
In plant imaging spectroscopy, we can utilize the interaction of solar radiation with plants. Generally, reflectance by single leaves or canopies is low in the visible spectrum (400–700 nm). This low reflectance is due to chlorophyll (the most abundant pigment in the plants), which have a characteristic reflectance peak in the green region of approximately 550 nm. However, when plants are analyzed using near infrared (NIR) wavelengths (700–1200 nm), there is a sharp increase in reflectance, because leaves reflect a large proportion of input radiation.
This principle has enabled the development of near-infrared, multispectral and hyperspectral imaging. Some of the applications using imaging spectroscopy include measurements of green biomass, canopy chlorophyll content, leaf and canopy senescence (or if they stay green) and plant water status.
Multispectral and hyperspectral imaging are widely used to estimate the canopy water content (as an indicator of water status), which uses the absorption bands in the infrared range to describe various water indices. Both differ in the number of bands. It means multispectral imaging encompasses 3-10 wider bands while hyperspectral has hundreds of narrow bands (GISGeography, 2021). A band is a segment of the electromagnetic spectrum. For instance, the visible spectrum has 380 to 700 nanometers in length and it has seven bands (or segments) corresponding to the colors the human eye can see to know red, orange, yellow, green, blue, indigo, and violet. Each of the bands in the visible spectrum has about 125 nm in length. In the case of hyperspectral imaging, these bands are reduced to 10-20 nm.
Plant imaging spectroscopy uses near-infrared instruments, spectrometers, hyperspectral and thermal cameras to capture the data of the plant reflectance. Furthermore, it allows for the analysis of different phenotype parameters such as leaf and canopy water status, leaf, and canopy health status, leaf growth, and coverage density.
There are some free public software and online tools which help to facilitate image analysis. For instance, the ImageJ software is very useful for different multi-scale image processing. Some of its applications include image registration, Landmark detection and multiscale color analysis and many more available at ImageJ website (Schneider et al 2012).
Furthermore, Quantitative Plant is another website that presents image analysis software tools and models for plants that are very user-friendly with access to image datasets and models (Lobet et al 2013).
Fang, Y., & Spector, D. L. (2010). Live cell imaging of plants. Cold Spring Harbor Protocols, 5(2), 1–5. https://doi.org/10.1101/pdb.top68
GIS Geography. (2021). Multispectral vs Hyperspectral Imagery Explained What’ s the difference between multispectral vs hyperspectral imagery? GISGeography. https://gisgeography.com/open-source-remote-sensin...
Higashiyama, T., Maizel, A., & Simon, R. (2021). Seeing is Believing: Advances in Plant Imaging Technologies. Plant and Cell Physiology, 62(8), 1217–1220. https://doi.org/10.1093/pcp/pcab133
Li, L., Zhang, Q., & Huang, D. (2014). A review of imaging techniques for plant phenotyping. Sensors (Switzerland), 14(11), 20078–20111. https://doi.org/10.3390/s141120078
Lobet, G., Draye, X., & Périlleux, C. (2013). An online database for plant image analysis software tools. Plant Methods, 9(1). https://doi.org/10.1186/1746-4811-9-38
Nguyen, V. D., Sarić, R., Burge, T., Berkowitz, O., Trtilek, M., Whelan, J., Lewsey, M. G., & Čustović, E. (2021). Noninvasive imaging technologies in plant phenotyping. Trends in Plant Science, xx(xx), 9–10. https://doi.org/10.1016/j.tplants.2021.06.009
Rousseau, D., Chéné, Y., Belin, E., Semaan, G., Trigui, G., Boudehri, K., Franconi, F., & Chapeau-Blondeau, F. (2015). Multiscale imaging of plants: Current approaches and challenges. Plant Methods, 11(1), 1–9. https://doi.org/10.1186/s13007-015-0050-1
Schneider, C. A., Rasband, W. S., & Eliceiri, K. W. (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods, 9(7), 671–675. https://doi.org/10.1038/nmeth.2089
Valuchova, S., Mikulkova, P., Pecinkova, J., Klimova, J., Krumnikl, M., Bainar, P., Heckmann, S., Tomancak, P., & Riha, K. (2020). Imaging plant germline differentiation within arabidopsis flowers by light sheet microscopy. ELife, 9, 1–19. https://doi.org/10.7554/eLife.52546
Walter, A., Liebisch, F., & Hund, A. (2015). Plant phenotyping: From bean weighing to image analysis. Plant Methods, 11(1), 1–11. https://doi.org/10.1186/s13007-015-0056-8