Skip to main content
ARS Home » Southeast Area » Stuttgart, Arkansas » Dale Bumpers National Rice Research Center » Research » Publications at this Location » Publication #276370

Title: A digital photography and analysis system for estimation of root and shoot development in rice weed suppression studies in the field

item Gealy, David
item Pinson, Shannon

Submitted to: Rice Technical Working Group Meeting Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: 12/19/2011
Publication Date: 2/1/2013
Citation: Gealy, D.R., Pinson, S.R. 2013. A digital photography and analysis system for estimation of root and shoot development in rice weed suppression studies in the field. 35th Rice Technical Working Group Meeting Proceedings. February 27, - March 1, 2012, Hot Springs, Arkansas. CDROM

Interpretive Summary:

Technical Abstract: Rice germplasm with an inherent ability to suppress weeds can potentially improve the economics and sustainability of weed control in rice. We devised a simple, rapid, and inexpensive digital imaging system to quantify several shoot and root growth characteristics in field-grown rice plants that have been associated with weed suppression potential. One possible use for the system will be to analyze large numbers of field-grown plants as would be required for the study of genetic mapping populations or other large plant populations. In order to field-test the system, we evaluated selected lines from a Lemont X Teqing ("Teqing into Lemont" or "TIL") mapping population in which the lines differed in weed suppression activity and growth habit. Up to 13 siblines were grown in drill-seeded plots in Stuttgart, AR in 2009, 2010, and 2011, and replicated 4 times. Four rice plants were randomly sub-sampled from each plot 7 to 10 days after establishment of the permanent flood when plants were approximately 60 cm tall. Roots were removed from the plots by maintaining them intact in a large plug of soil to minimize damage. They were then immersed in water and rinsed to remove soil, and the entire intact plant was laid on a printed background grid, and photographed on a platform at the field site using a digital camera (Canon Power Shot 3G; 4.0 mega pixels; 7.2-28.8 mm zoom lens). Photographs were taken using a fixed camera mount under a rigid frame enclosed in opaque white sheets to prevent interference from shadows and glare. The background grid was white with calibrated black lines projecting upwardly and downwardly at several fixed angles from the point at which the base of the plant shoot was placed. Lengths and growth angles of shoots (stems + leaves) and roots, and tiller numbers could be visually estimated directly from the photographs. Areas of roots and shoots were estimated from the digital photographs using commercial software (Adobe Photoshop Elements ver. 9). By single-clicking the mouse within the "brown-colored" root area or the "green-colored" shoot area, the software provided a total pixel count (relative area) for both roots and shoots. Several standardized squares of known area (100 cm**2; colored red) were printed near the edges at the top, middle, and bottom of the background grid where they were unobstructed by plant tissues, so that pixel counts could be accurately converted to true areas (cm**2). Preliminary analyses showed that digital area estimates for both roots and shoots usually correlated with tissue dry weights. We presented a digital imaging system for rapid estimation of plant root and shoot growth, and tillering patterns of rice lines using inexpensive software. Potential advantages of the system are that numerous plant parameters from field-grown plants can be estimated rapidly, and that the digital photographic images can be analyzed indoors under more comfortable conditions, and at a later time, which helps space out the work load during the busy season. The system appears to be readily adaptable for similar analyses of barnyardgrass plants or other weeds from rice fields. Plant-to-plant variability within plots may reduce the effectiveness of the system, requiring multiple subsamples for best results.