Crop Phenotyping of Sorghum bicolors Physiological Response to Salt-Affected Soils Using TLS and GPR Remote Sensing Technologies in Nevada Drylands
Loading...
Authors
Smith, Erin Louise
Issue Date
2023
Type
Thesis
Language
Keywords
Alternative Title
Abstract
Saline and sodic soils are major abiotic stressors on the production of flood-irrigated crops in drylands. We conducted a crop phenotyping, remote sensing study on five genotypes of sorghum [Sorghum bicolor (L.) Moench], a drought and salt-tolerant crop, to assist in the molecular breeding of salt-tolerant cultivars. A control plot and a spatially heterogeneous saline-sodic plot (treatment plot) were established in collaboration with Dr. Yerka, Mr. Alfredo Delgado, Dr. Washington-Allen, the Nevada Agricultural Experiment Station (NAES) and the United States Department of Agriculture’s Plant Materials Center (USDA-PMC) in Fallon, Nevada. This location is representative of the variable salinity/sodicity conditions typical of Northern Nevada soils and associated belowground biomass dynamics in drylands. We generated pre- and post-harvest soil attribute maps of the treatment plot using spatial interpolation, we expected individual genotypes to be affected differently by the gradient of various soil constituents. We hypothesized that above- and belowground three-dimensional structural phenology of the five genotypes would be differently affected across the salinity gradient in the treatment plot relative to the control plot. Additionally, we hypothesized that the GPR signal return would vary with the salinity gradient. Finally, we expected an increase in belowground biomass, relative to the control plot, in response to salt-stress as an adaptation to drought. The phenology of coarse-root depth and three-dimensional structure from pre-planting to harvest was non-invasively measured 15 times using a real-time kinematic (RTK) GPS-mounted IDS GeoRadar dual channel (400MHz and 900MHz) ground penetrating radar (GPR) system. Plant height and three-dimensional structural phenology of the five varieties were mapped using a FARO Focus3D X 330 terrestrial laser scanner (TLS). We found differences in above- and belowground three-dimensional structural phenology across the five genotypes in response to the salinity and sodicity gradient. Of the five genotypes in this study, only four emerged in the treatment plot, where Richardson Seed’s Ultra-Early Hybrid performed best under the gradient of salinity and sodicity with the highest rate of emergence (68%), the highest rate of panicle production (4.1 panicles per row), and the greatest panicle volume (67.2%) relative to the control plot. Furthermore, we found that the GPR return signal was not able to detect root mass in the highly saline-sodic soil, however, I was able to detect root mass phenology in the control plot. GPR return signal was not linear in response to the salinity gradient, however, a signal pattern emerged from the different salinity ranges suggesting a gradient response. This study shows the efficacy of the use of these technologies in crop phenotyping and precision agriculture. Future work may improve TLS derived data processing efficiency by developing methods for automating the detection of phenotypic traits (e.g., panicles, leaf area index, number of individual plants). These methods likely will include machine learning algorithms, allometric equations for biomass calculations, and use of drone-mounted LiDAR to reduce occlusion. The use of GPR in the salt-affected soils of this study was not able to definitively identify root mass, however, its use in soil composition for salts and other constituents is indeed promising. Further testing of GPR’s non-detect threshold in salt-affected soils and its ability to quantify individual soil constituents has potential to be highly valuable to the field of soil science and precision agriculture. Furthermore, this study was able to detect a root mass response using GPR, future work may focus on differentiating genotypic variation in root phenology.