2017 ASHS Annual Conference

Monitoring the growth of bedding plant seedlings is essential to ensure quality production. However, visual growth assessments are often erroneous, while manual measurements can be time consuming due to the high density of seedlings in a plug tray. A strong correlation between leaf area and the accumulation of biomass exists for many plant species. Generally, as leaf area increases the amount of light intercepted by the plant also increases. We tested the use of automated and non-invasive imaging using fluorescence to estimate leaf area and track the growth rate of entire plug trays. Our hypothesis was that the fluorescence from seedlings could be effectively used to separate plant material from the surrounding area during image processing, thereby increasing the accuracy of measurements. Seeds of pansy (Viola ×wittrockiana ‘Matrix^TM Yellow’), petunia (Petunia ×hybrida ‘Dreams Midnight’), tomato (Lycopersicum esculentum ‘Early Girl’), and zinnia (Zinnia elegans ‘Zahara Fire’) were sown in 128-cell plug trays and grown in a greenhouse environment under an average daily light integral of 16.1 ± 3.5 mol·m^–2·d^‒1 and an average daily temperature of 20.1 ± 0.67 °C. Data collection occurred every two days starting from the third day after germination, with one tray from each species randomly selected for imaging and destructive measurements. Images were collected after exposing seedlings to blue light (470 nm) and filtering fluorescence from the seedlings using a top view image station (Aris BV, Zeelsterstraat 80, The Netherlands). The image processing was rapid and automatic after collection using OpenCV software to estimate the pixel area. Destructive data collection of each tray for leaf area, root dry mass, and shoot dry mass immediately followed imaging. Regression analyses indicated a strong linear relationship (r² = 0.95 to 0.99) between destructively measured leaf area and non-invasive pixel area from fluorescence imaging, shoot dry mass, and total (root + shoot) dry mass for all species. Therefore, the proposed imaging technique could be utilized to estimate the leaf area and total biomass of entire plug trays, with potential use in the bedding plant industry as a means to efficiently and non-destructively monitor the growth rate of bedding plants seedlings.