Author
LUO, WEIQI - North Carolina State University | |
Gottwald, Timothy | |
POSNY, DREW - North Carolina State University |
Submitted to: Journal of Citrus Pathology
Publication Type: Abstract Only Publication Acceptance Date: 12/20/2016 Publication Date: 5/18/2017 Citation: Luo, W., Gottwald, T.R., Posny, D. 2017. Calculating historical citrus reduction rate/pattern using aerial photographic and GIS techniques. Journal of Citrus Pathology. 4(1):26/45. Interpretive Summary: Aerial photographs provide snapshots of spatio-temporal disease progression through canopy reduction and tree removal. By using citrus field imagery from 2006 to 2015, we investigate the spread of HLB through different landscapes. Identifying how the disease progressed in the past can elucidate unexpected patterns, risks, and spread timeframes that can provide unique guidance for future HLB survey methodologies. Technical Abstract: Ever since first detection in late 2005, HLB is now widely spread throughout most commercial citrus groves in Florida and caused a rapid decline of trees, as the disease continues to tighten its grip on Florida's citrus industry. However, there are limited studies that quantify the impact of HLB on citrus reduction rate over large spatial scales. Aerial photographs are most commonly used to measure historical changes since the advancement and availability of high-resolution, large-scale vertical aerial photography. This study presents a semi-automatic statistical classification method for annual citrus tree decline rate calculation and pattern recognition/analysis. High-resolution photography of Florida (~1 foot) from 2006 to 2015 can be obtained from LABINS (Land Boundary Information System), where citrus field imagery is converted to digital format and filtered out using user-defined citrus polygons in GIS. Subjected to selection bias, a number of commercial citrus fields are used to fully represent the spatial variability of the HLB situation. In order to estimate active citrus coverage accurately, we specified multiple training areas as the color-spectral properties of citrus canopy photography are apt to change for different times and locations. Also, efforts have been made to remove tree shadows by identifying neighboring points with similar color intensity. Transition from healthy to serious infected citrus grove results in potential significant changes in tree canopy size (e.g. most likely due to canopy decline/shrinkage and diseased tree removal). We compared annual tree coverage reduction rates across different landscapes, identified major patterns of tree loss within grove (i.e. edge effect), and quantified the relationship with HLB progress in Florida. Incorporating both spatio-temporal analysis and aerial photography has the benefit in determining the economic impact of HLB on actual citrus landscape. This proposed technique could be used as an enhancement tool in future HLB survey design and modeling. |