| Abstract Detail
Development and Structure Thrash, Tyler [1], Lee, Hansol [2], Gerbitz, Ryan [2], Kozyra, Abigail [3], Baker, Robert (Rob) [4]. A low-cost high-throughput phenotyping system for studying developmental trends and abiotic stress. Abiotic stress is a critical consideration for most crop improvement programs and may be complicated by developmental changes in stress tolerance. For example, salt stress can affect different organs of Brassica rapa at different periods of development. However, studies on the development of plant morphology in response to abiotic stress are often limited with respect to cost, scalability, and the temporal density of measurements. Automated measurements with high-throughput phenotyping (HTP) systems can allow for a higher spatial resolution and a higher temporal density than measurements that are acquired manually. However, HTP systems are often expensive and cannot easily scale to different environments. For the present study, we developed and validated an inexpensive and scalable HTP system using eight cameras and a Raspberry Pi computer. The cameras were installed over greenhouse benches during an experiment and controlled wirelessly and automatically by the PI computer for over a month. RBG images were acquired from each camera approximately once every hour for the entire duration of the experiment with minimal intervention. Throughout the experiment, four benches containing up to 70 plants (Brassica rapa) were watered to field capacity with either tap water or a 0.3% NaCl solution. On five different days early in development, we also manually acquired high-resolution images of each individual plant under controlled lighting conditions for comparison with the HTP images. After the experiment, a “greenness index’ was computed for each pixel of both image types and was used to estimate foliar size and overall greenness. We found that the HTP system reproduced developmental trends and treatment differences compared to the manually acquired images. In addition, our low-cost HTP system allowed us to acquire these estimates for a long period of time, at a high temporal density, under varying lighting conditions, and with minimal manual intervention. Future work will focus on the application of this HTP system for detecting genotype-by-environment interactions over development and the scalability of this system for larger field studies.
1 - Saint Louis University, Biology, 3507 Laclede Avenue, MWN 317, St. Louis, MO, 63103, United States
2 - Miami University, Biology, 212 Pearson Hall, Oxford, OH, 45056, USA
3 - Miami University, 212 Pearson Hall, Oxford, OH, 45056, USA
4 - National Park Service, Inventory And Monitoring Division, 1201 Oakridge Drive, Suite 150, Fort Collins, CO, 80525, United States
Keywords:
agriculture technology
image analysis
controlled environment.
Presentation Type: Oral Paper
Number: DS3001
Abstract ID:129
Candidate for Awards:Katherine Esau Award,Developmental and Structural Section Graduate Student Registration Award |
