Sugarcane Yield, Sugarcane Quality, and Soil Variability in Louisiana

Authors: Johnson, Richard; Richard Jr, Edward

Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/16/2004
Publication Date: 5/20/2005
Citation: Johnson, R.M., Richard Jr, E.P. 2005. Sugarcane Yield, Sugarcane Quality, and Soil Variability in Louisiana. Agronomy Journal. 97:760-771.

Interpretive Summary: Louisiana sugarcane producers, like most U.S. agricultural producers, have faced increased economic pressure in recent years. Growers must find ways to decrease costs and maximize profits. Precision agriculture has particular potential for increasing profitability while decreasing costly variations in sugar quality, variations that decrease both the price paid to the producer and the cost of processing the sugar at the mill. In this study, soil analyses, an important tool in precision agriculture, was combined with sugar analysis to map the correlations among soil fertility, sugarcane quality, and yield. Distinct relationships were found between soil properties and yield. The yield correlations were particularly significant with respect to soil organic matter, soil calcium, and soil magnesium; and those soil properties, in addition to soil sulfur and soil potassium, were strongly linked to sugar quality. These properties were also found to vary across the field. The observed variation was not random, but was instead found to be spatially correlated. 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. To further study this variation, precision agriculture maps were constructed for each soil and sugarcane property. A strong relation was found between soil potassium, soil organic matter, soil pH, soil calcium and soil sulfur and sugarcane yield and quality. These maps can also be used to develop variable rate fertilizer and lime programs. Sugarcane growers that adopt these methods may benefit from decreased fertilizer and lime costs, while minimizing the environmental impact of their application.

Technical Abstract: This study reports results from our evaluations of the spatial variability of sugarcane (interspecific hybrids of Saccharum spp.) yield and quality in relation to soil chemical properties in Southern Louisiana. Sugarcane cv 'LCP 85-384' was grown in two locations, Rebecca Plantation at Schriever, LA and Gralyn Farms at Patoutville, LA. Each field was harvested in a grid cell pattern with cell dimensions of 10.6 x 15.2 m. Individual cells were harvested with a single row, chopper sugarcane harvester with weights of the harvested stalk pieces (billets) determined using a field transport wagon equipped with electronic load sensors. In addition to weights, a sample of the harvested cane was randomly obtained from each grid cell for sugar quality analysis and soil samples (0-20 cm) were collected after harvest from each grid cell. Soil properties determined included: OM, pH, Ca, Mg, K, P, CEC, and S. Yield, quality, and soils data were analyzed by both conventional statistics and geostatistical techniques. At Rebecca Plantation, the majority of soil properties exhibited non-normal distributions with coefficients of variation ranging from 1 to 47% in 2002 and from 1 to 50% in 2003. All soil properties were spatially correlated with the range of spatial correlation varying from 29 to 138 m in 2002 and from 71 to 241 m in 2003. At Gralyn Farms, the majority of soil properties also exhibited non-normal distributions with coefficients of variation ranging from 1.5 to 50% in 2002 and from 1 to 56% in 2003. All soil properties were spatially correlated with the range of spatial correlation varying from 26 to 192 m in 2002 and from 95 to 166 m in 2003. The majority of properties describing sugar yield and quality at both Rebecca Plantation and Gralyn Farms were found to possess non-normal distributions in 2001. This number decreased in 2002 and 2003. The coefficients of variation were relatively stable between location and year varying from 5 to 20%. At Rebecca Plantation in 2001, 2002, and 2003, all sugar yield and quality parameters were spatially correlated with the exception of TRS and Fiber in 2003. The ranges of spatial correlation varied from 26 to 187 m. At Gralyn Farms in 2001, 2002, and 2003, all sugar properties were spatially correlated with ranges of spatial correlation varying from 27 to 133 m. Correlation analysis was performed to study the relation between soil properties and sugar parameters. At Rebecca Plantation, the soil Ca:Mg ratio and soil S were correlated to all sugar parameters and at Gralyn Farms, soil OM, and soil pH were correlated to all sugar parameters. Multiple regression equations were developed with these same properties to describe selected sugar parameters at both locations. Kriged maps of sugarcane yield, quality and soil properties illustrated the distinct relations between soil variability and sugar yield and quality. The combined data suggest that these fields could benefit from a precision management approach.