Developing the framework for a risk map for mite vectored viruses in wheat resulting from pre-harvest hail damage

Abstract

There is a strong economic incentive to reduce mite-vectored virus outbreaks. Most outbreaks in the central High Plains of the United States occur in the presence of volunteer wheat that emerges before harvest as a result of hail storms. This study provides a conceptual framework for developing a risk map for wheat diseases caused by mite-vectored viruses based on preharvest hail events. Traditional methods that use NDVI were found to be unsuitable due to low chlorophyll content in wheat at harvest. Site-level hyperspectral reflectance from mechanically hailed wheat increased canopy albedo. Therefore, any increase in NIR combined with large increases in red reflectance near harvest can be used to assign some level of risk. The regional model presented in this study utilized Landsat TM/ETM+ data and MODIS imagery to help gapfill missing data. NOAA hail maps that estimate hail size were used to refine the area most likely at risk. The date range for each year was shifted to account for annual variations in crop phenology based on USDA Agriculture statistics for percent harvest of wheat. Between 2003 and 2013, there was a moderate trend (R 2 = 0.72) between the county-level insurance claims for Cheyenne County, Nebraska and the area determined to be at risk by the model (excluding the NOAA hail size product due to limited availability) when years with low hail claims (< 400 ha) were excluded. These results demonstrate the potential of an operational risk map for mitevectored viruses due to pre-season hail events.