Techniques for determining disease variability and yield loss in commercial sugarcane

Authors: Grisham, Michael; Johnson, Richard; Viator, Ryan; Zimba, Paul

Submitted to: International Society of Sugar Cane Technologists Pathology Workshop
Publication Type: Abstract Only
Publication Acceptance Date: 3/27/2008
Publication Date: 6/23/2008
Citation: Grisham, M.P., Johnson, R.M., Viator, R.P., Zimba, P.V. 2008. Techniques for determining disease variability and yield loss in commercial sugarcane [abstract]. Proceeding of International Society of Sugar Cane Technologists Pathology Workshop. Presentation No. 29.

Interpretive Summary:

Technical Abstract: Researchers frequently use replicated, small-plot experiments to measure differences in yield between infected and diseased plants. To provide a better understanding how environmental conditions and cultural practices may influence the incidence and severity of disease outbreaks, we are conducting research in commercial sugarcane fields that incorporates techniques commonly used in the study and application of precision agriculture. We used these techniques to study the incidence and severity of brown rust (Puccinia melanocephala) infestations in five commercial fields. A handheld computer equipped with a GPS (global positioning system) receiver and mapping software was used to determine experimental field boundaries and establish grid-sampling points. Field size varied from 0.7 to 8.9 hectares, and grid size varied from 0.02 to 0.4 hectares. At each grid-sampling point, soil samples were taken and weekly rust ratings were for six or seven weeks. Experimental plots were harvested green utilizing a single-row, chopper harvester and the total weight of harvested cane in each plot was determined using a single-axle, high-dump billet wagon containing three electronic load sensors and equipped with a device to collect a billet sub-sample from each plot for sucrose quality analysis. Soil properties, rust ratings, and cane yield varied significantly among the sampling points within fields and also between fields. In addition, overall fertility levels of each location were variable. Rust ratings were found to be positively correlated with several soil properties, most notably phosphorus and sulfur. These correlations were even more pronounced when phosphorus and sulfur was present at excessive levels. Sucrose and cane yields were negatively correlated to the rust ratings, with higher rust levels causing reduced yields even in areas of excess nutrition. These combined data suggest that sugarcane growers that apply fertilizer at rates exceeding plant requirements will increase the incidence and severity of rust infestations in their fields. Remote sensing is another tool used in precision agriculture research that we are applying to sugarcane pathology research. We compared reflectance of leaves infected with either Sugarcane yellow leaf virus (SCYLV) or Sorghum mosaic virus (SrMV) to leaves from noninfected plants using a fiber optic spectrometer mounted 3 cm above the leaf. From analysis of leaf reflectance measurements, leaves exhibiting either mild or severe mosaic symptoms were correctly classified in 75 and 68% of the cases, respectively; and leaves infected with SCYLV were correctly identified 77% of the time even though no visible symptom of yellow leaf was expressed. Remote sensing provides an alternative method of disease diagnosis and detection that offers a practical and economical method of monitoring disease development.