Relationship between sugarcane rust severity and soil properties in Louisiana

Authors: Johnson, Richard; Grisham, Michael; Richard Jr, Edward

Submitted to: Phytopathology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/9/2006
Publication Date: 6/1/2007
Citation: Johnson, R.M., Grisham, M.P., Richard Jr, E.P. 2007. Relationship between sugarcane rust severity and soil properties in Louisiana. Phytopathology. 97:748-755.

Interpretive Summary: Brown rust of sugarcane was not considered to be of major importance in Louisiana until 2000 when an epidemic occurred throughout the sugarcane industry. The outbreak and its continued presence is of concern because the most severely affected variety is LCP 85-384, a variety that occupies the majority of the sugarcane acreage in Louisiana. A total of five sugarcane fields were sampled in a grid pattern to determine if a relation could be found between soil properties and sugarcane rust levels. At each grid point within a field, soil samples were collected and then rust severity was monitored at weekly intervals for a period of six to seven weeks. The soil properties at each location showed significant variability with coefficients of variation ranging from 9 to 70% over all locations. Soil phosphorus exhibited the greatest degree of variability, followed by soil magnesium, soil calcium and soil sulfur. In addition, soil properties were spatially correlated in 39 of 40 cases and rust ratings were spatially correlated in 32 of 33 cases. In other words, samples that were spaced at a distance less than the range of spatial correlation were more similar than those at distances greater than this range. The range of spatial correlation for soil properties varied from 39 to 201-meters and from 29 to 241-meters for rust ratings. Rust ratings were correlated with several soil properties, most notably soil phosphorus and soil sulfur. Other soil properties correlated with rust occurrence include; soil pH, soil potassium, soil calcium, soil magnesium, soil organic matter and soil cation exchange capacity. When all locations were combined the best correlations were obtained with soil sulfur, phosphorus, and soil pH. Taken together, these relations indicate that rust severity increased with soil fertility levels. This also suggests that sugarcane growers that apply fertilizer in excess of plant requirements will increase the incidence and severity of rust infestations in their fields.

Technical Abstract: This study was conducted to determine the extent of temporal and spatial variability of sugarcane rust (Puccinia melanocephala) present in commercially cultivated sugarcane (interspecific hybrids of Saccharum spp., cv ‘LCP 85-384’) grown in South Louisiana. Sugarcane fields at two locations in Gheens, LA and three locations in Schriever, LA were grid-mapped at several intensities ranging from one sample per 0.05 to 0.5-hectares. Soil samples (0-15-cm) were collected at each grid point and the rust severity was monitored at each point at weekly intervals for a period of six to seven weeks. The soil properties at each location exhibited significant variability with coefficients of variation ranging from 9 to 70.1% over all locations. Soil phosphorus exhibited the greatest degree of variability, followed by soil magnesium, soil calcium and soil sulfur. In addition, soil properties were spatially correlated in 39 of 40 cases and rust ratings were spatially correlated in 32 of 33 cases. The range of spatial correlation for soil properties varied from 39 to 201-meters and from 29 to 241-meters for rust ratings. Rust ratings were correlated with several soil properties, most notably soil phosphorus (r = 0.40 to 0.81) and soil sulfur (r = 0.36 to 0.68). Other soil properties correlated with rust occurrence at selected locations and times include; soil pH, soil potassium, soil calcium, soil magnesium, soil organic matter and soil cation exchange capacity. When all locations were combined the best correlations were obtained with soil sulfur, phosphorus, and soil pH. Multiple linear regression models constructed from soils data to predict initial rust infestation levels resulted in coefficients of determination that ranged from 0.22 to 0.73. Models were most effective at predicting rust severity within a location. Discriminant analysis of soil data significantly improved the overall predictive ability of rust models. Finally, contour plots of soil properties and rust levels provide further evidence of a link between these two parameters. Taken together, these relations indicate that rust severity increased with soil fertility levels. This also suggests that sugarcane growers that apply fertilizer in excess of plant requirements will increase the incidence and severity of rust infestations in their fields.