Modeling the relationship between southern highbush blueberry flower density and flower thrips (Frankliniella spp.) abundance

Southern highbush blueberries are a highly profitable crop in Florida. Flower thrips (Frankliniella spp.) are the key pests of these blueberries. They feed and oviposit in all flower structures. The injury caused by these activities can lead to fruit scarring. Over time, thrips population density is highly correlated with flower phenology. The hypothesis of this paper was that thrips density is also correlated with flower density in space. In order to test this hypothesis, 130 white sticky traps were placed in a blueberry planting in Inverness, Florida during early spring 2009. Traps were changed out weekly until fruit set began. Percent of open flowers was recorded from each sampling row when the traps were replaced. That number was then assigned to each sample point in the row. Since the data did not meet the assumptions for least squares regression, Theil regression was used to analyze it. ArcGIS was used to create maps showing where in the field like (low with low or high with high) and unlike values of percent opened flowers and thrips per trap were paired. There was a positive linear relationship between percent opened flowers and thrips per trap during the 1st, 2nd, and 4th sampling weeks. Peek flowering occurred during the 3rd sampling week and the thrips population remained high during the 5th sampling week when fruit set was beginning. This may account for the lack of a relationship between flower density and thrips abundance during these weeks. These trends were corroborated by the GIS maps.