Spatial and Temporal Physiognomies of Whitefly and Tomato Yellow Leaf Curl Virus Epidemics in Southwestern Florida Tomato Fields

Abstract

Epidemics of tomato yellow leaf curl virus (TYLCV; species Tomato yellow leaf curl begomovirus) have been problematic to tomato production in the southeastern United States since the first detection of the virus in Florida in the late 1990s. Current strategies for management focus on farm-centric tactics that have had limited success for controlling either TYLCV or its whitefly vector. Areawide pest management (AWPM)—loosely defined as a coordinated effort to implement management strategies on a regional scale—may be a viable management alternative. A prerequisite for development of an AWPM program is an understanding of the spatial and temporal dynamics of the target pathogen and pest populations. The objective of this study was to characterize populations of whitefly and TYLCV in commercial tomato production fields in southwestern Florida and utilize this information to develop predictors of whitefly density and TYLCV disease incidence as a function of environmental and geographical factors. Scouting reports were submitted by cooperating growers located across approximately 20,000 acres in southwestern Florida from 2006 to 2012. Daily weather data were obtained from several local weather stations. Moran’s I was used to assess spatial relationships and polynomial distributed lag regression was used to determine the relationship between weather variables, whitefly, and TYLCV. Analyses showed that the incidence of TYLCV increased proportionally with mean whitefly density as the season progressed. Nearest-neighbor analyses showed a strong linear relationship between the logarithms of whitefly densities in neighboring fields. A similar relationship was found with TYLCV incidences. Correlograms based on Moran’s I showed that these relationships extended beyond neighboring fields and out to approximately 2.5 km for TYLCV and up to 5 km for whitefly, and that values of I were generally higher during the latter half of the production season for TYLCV. Weather was better at predicting whitefly density than at predicting TYLCV incidence. Whitefly density was best predicted by the number of days with an average temperature between 16 and 24°C (T16to24), relative humidity (RH) over the previous 31 days, and vapor pressure deficit over the last 8 days. TYLCV incidence was best predicted by T16to24, RH, and maximum wind speed over the previous 31 days. Results of this study helped to identify the extent to which populations of whitefly and TYLCV exist over the agricultural landscape of southwestern Florida, and the environmental conditions that favor epidemic growth. This information was used to propose an approach to AWPM for timing control measures for managing TYLCV epidemics.