Color preference, seasonality, spatial distribution and species composition of thrips (Thysanoptera: Thripidae) in northern highbush blueberries

We investigated color preference, seasonal abundance, spatial distribution and species composition of thrips in northern highbush blueberries, Vaccinium corymbosum L., in New Jersey (USA). White sticky traps were more attractive to thrips compared with yellow or blue traps. Thrips captures using white sticky traps showed that their flight activity begins 20e30 d after the onset of flowering, with 10, 50 and 90% of trap captures observed at 383, 647 and 1231 degree-day accumulations, respectively (10 C base temperature). Two methods were used to study thrips distribution within a blueberry bush. First, white sticky traps were placed within the bush canopy at three different heights. The highest numbers of thrips were caught on traps in the middle and top one-third of the canopy while the lowest numbers were caught in the bottom one-third. A second method determined the distribution of thrips on the blueberry plant at different heights and phenological stages. The highest numbers of thrips were found on young leaves at lower parts of the canopy, whereas flowers and fruit had fewer thrips and none were found on buds; these thrips were identified as, Scirtothrips ruthveni (88% of adults) and Frankliniella tritici (12%). The distribution of thrips within a blueberry planting was investigated using an evenly-spaced grid of white sticky traps in combination with on bush beating-tray samples. Thrips counts from traps correlated with direct counts on the bush across the entire blueberry field (macro-scale level); however, within the field (micro-scale level), there was no correlation between the number of thrips on traps and on individual bushes near traps. Early in the season, trap counts were higher on bushes closer to the forest, indicative of movement of thrips fromwild hosts into blueberry fields. However, this was not the case for direct on bush counts or trap counts for the later part of the season, where there was no clear forest “edge” effect. Percent fruit injury due to thrips feeding was low, and it correlated with thrips counts on bushes but not from counts on traps. Overall, our data show that thrips counts on sticky traps need to be interpreted with care because these numbers weakly correlated with the numbers of thrips on bushes at the micro-scale level and percent fruit injury; however, they can be useful predictors of thrips activity across entire blueberry fields (macro-scale). 2010 Elsevier Ltd. All rights reserved.