High-throughput Image-based Phenotyping for Studying Plant Respo to Arbuscular Mycorrhizal Fungi

摘要

Arbuscular mycorrhizal fungi (AMF), as plant bio-stimulants, have been gaining prominence in agriculture because of their potential in improving nutrient uptake efficiency, tolerance to various stressors, and crop growth (Rouphael, Spichal, Panzarova,et al. 2018). Multiple benefits to plants by AMF include uptake of phosphorus and other poorly mobile nutrients (for example, N, S, K, Ca, Fe, Cu, Zn), increased soil aggregation and carbon sequestration, and stimulation of other organismal growth in the rhizosphere (Ingraffia, Amato, Frenda, et al. 2019; Battini, Gronlund, Agnolucci, et al.2Ql). Responses to the mycorrhizal colonization may vary widely, depending on the host plant as well as the fungus. The aforementioned responses are positive for plant's growth, but there are a few negative responses exhibited by plants when exposed to AMF, for example, mycorrhizal growth depressions.These negative responses are gaining popularity amongst researchers/scientists engaged in AMF development. These responses can be observed in the form of phenotypic changes in plants. Plants respond to AMF and are colonized by them at different times and intensities. A better understanding of plant growth response to mycorrhizae can be achieved by collecting time series data of the plant phenotype. Such data sets can be laborious and costly to generate, as they usually rely on multiple harvests (Watts-Williams, Jewell, Brien, et al. 2019). An emerging advancement in obtaining such data is the use of image capture technology and open source analysis tools (Fahlgren, Gehan, and Baxter 2015). Plant phenotyping is an important tool in addressing and understanding plant environment interaction and its translation into application in crop management practices and effectsof biostimulants (Pieruschka and Schurr 2019). Plant phenomics employs high-throughput image-based phenotyping of above-ground plant tissues (Marko, Brigila, Summerer, et al. 2018; Fahlgren, Gehan, and Baxter 2015). 'High-throughput' is a technologicalclassification term that is relative to the effort associated with measurement (Fahlgren, Gehan, and Baxter 2015). Improvements in the specificity and throughput of phenotypic assessment at different biological levels such as phenomics, metabolomics, andgenomics, are the main objectives of modernphenotyping (Sytar, Zivcak, Olsovska, etal. 2018). Modern phenotyping comprehensively assesses complex plant traits such as growth, development, tolerance, resistance to biotic and abiotic stresses, architecture, physiology and yield, through automated processes (Marko, Brigila, Summerer, etal. 2018; Fahlgren, Gehan, and Baxter 2015). Recent advances in this technology have enabled rapid worldwide development of high-throughput automated and semi-automated field and laboratory phenotyping platforms. Thenewly developed plant phenotyping platforms produce relatively large quantity of data than the initial platforms; hence need special systems for data management, access and storage, and also new statistical tools for deriving biologically significant signals from both experimental and environmental noise (Sytar, Zivcak, Olsovska, et al. 2018).