Investigate the ecology and management of blueberry flower thrips in Georgia

Specific objective 1: Determine the spatio-temporal distribution pattern and alternative host of blueberry flower thrips in Georgia

Spatio-temporal distribution:

2.1 Study area and survey design

Flower thrips were collected from three southern highbush and three rabbiteye blueberry fields across three counties: Bacon, Appling, and Pierce. All six blueberry fields were located on commercial farms and were well-maintained by the growers, following the necessary agronomic practices. No insecticide applications on these farms were applied specifically for thrips management. Twenty sampling spots were selected in each field. The distance between the sampling spots was approximately 15-20 meters. Sampling spots were selected based on their distance from the field edge closer to the wild habitat, with four spots at the edge of the field and the remaining spots located at 15-meter, 30-meter, 45-meter, and 60-meter distances from the field edge towards the center of the field. GPS coordinates of each sampling point were recorded using iPhone 11. A double-sided, 3" x 5" yellow sticky card (Catchmaster brand) was placed in the bushes at each sampling location to monitor thrips populations. Frankliniella tritici was most attracted to yellow compared with blue or white traps in tomatoes (Cho et al. 1995). However, the yellow sticky card alone is not suitable for determining the flower thrips' population density. Therefore, flower samples were also collected randomly from each spot. 10 flower clusters were collected to determine the thrips abundance. Adult flower thrips were gathered from flower samples as well as yellow sticky cards. However, when it came to counting the population of larvae, only flower thrips samples were considered. Flower clusters from each spot was placed in the zip lock bags, marked with location, types of blueberries (southern highbush and rabbiteye), date, and spot number, and transported to the fruit entomology lab, Athens, Georgia, for processing. Similarly, weekly monitoring of thrips was done using yellow sticky cards during the blueberry season. Every week, yellow sticky cards were replaced with new yellow sticky cards.

2.2 Sample processing

The samples were brought to the lab where they were placed in a liter size plastic cup with 75% ethanol for 30 minutes to dislodge various life stages of thrips. Then, the flowers were taken out of the cup, leaving thrips in alcohol. The contents in alcohol were sieved using a 250-μm grating, USA Standard Testing Sieve (W. S. Tyler, Inc.). The residue in the sieve was washed off with 75% alcohol into a Petri dish and checked under a dissecting microscope at 12X to count the number of various thrips species. Subsamples of thrips were sent to Dr. Cheryle A. O’Donnell at the Systematic Entomology Laboratory (USDA APHIS PPQ in Beltsville) for accurate species identification.

Alternative host identification:

Identification of alternative host plants for flower thrips were recorded in adjacent wild habitats surrounding the six study sites, which included four commercial blueberry fields and two wild habitats in Bacon County, Georgia. Flower samples from dominant weed species were collected monthly from November to March. Sampling locations were systematically spaced 20 meters apart to ensure representative coverage and reduce spatial bias across the wild habitat. During the spring seasons of both years, flowers from the dominant weed species were specifically targeted to identify their role as alternative hosts for overwintering thrips. The selected weed species included wild radish (Raphanus raphanistrum, Brassicaceae), blue toadflax (Nuttallanthus canadensis, Plantaginaceae), dewberry (Rubus trivialis, Rosaceae), fetterbush (Lyonia lucida, Ericaceae), four-petal St. John’s wort (Hypericum tetrapetalum, Hypericaceae), heartwing sorrel (Rumex hastatulus, Polygonaceae), orange milkwort (Polygala lutea, Polygalaceae), spiderwort (Tradescantia ohiensis, Commelinaceae), wild blueberry (Vaccinium angustifolium, Ericaceae), and sand blueberry (Vaccinium myrsinites, Ericaceae). The sample size for each species was determined based on the size and availability of flowers from individual plants to ensure adequate representation for thrips assessment. Samples were carefully collected in containers filled with 70% ethanol to preserve the specimens. In the laboratory, each sample was examined under a stereo microscope, and all the thrips presents were identified, counted, and recorded. This study aimed to determine the role of common weed species in supporting overwintering thrips populations, offering valuable insights into potential alternative host plants that could serve as reservoirs for thrips during periods when blueberry flowers were not available.

Statistical analysis:

For spatio-temporal distribution, seasonal variation of F. tritici population including both adult and larva, role of blueberry types on thrips population and regression analysis to determine the role of maximum and minimum temperature on thrips population were analyzed and graphs were prepared using JMP pro software (SAS Institute Inc., Cary, NC, 1989-2023). In this study, the Inverse Distance Weighted (IDW) method was employed for interpolating the local aggregation indices, aiming to identify the clustering and uniform distribution patterns of flower thrips within blueberry fields. The analysis, using the SADIE method, allowed for the calculation of aggregation (cluster) and gap indices. SADIE analysis was conducted using the "epiphy" package in R version 3.6.3 (R Core Team 2020). Through SADIE analysis, local indices of dispersion were generated for each sampling point, indicating either a positive cluster index (Vi) for counts above the average or a negative gap index (Vj) for counts below the average.

To spatially interpolate the local aggregation or association indices, the inverse distance weighted (IDW) method was employed in ArcGIS Pro version 2.9. This method uses the inverse of the distances between sample points to assign weights for estimating values at unsampled locations. By applying the IDW method, a continuous surface representing the spatial distribution of the aggregation or association indices was created in ArcGIS Pro.

For alternative host analysis, differences in the mean number of adults recovered from weed species were analyzed using one-way ANOVA (P> 0.05; Tukey’s HSD test).

Specific objective 2: Characterization of flower thrips injury in blueberry

Experiments in Greenhouse:

This study was conducted in a greenhouse to quantify and characterize the injury caused by Frankliniella tritici to blueberry flowers and assess its effects on fruit development and yield. Four blueberry cultivars were used: two southern highbush (SHB) varieties (Farthing’ and ‘Suziblue’) and two rabbiteye (REB) varieties (Brightwell’ and ‘Climax’). Each plant was individually placed in a thrips-proof cage with a mesh size ≤ 150 µm (0.15 mm) to prevent thrips escape. For each plant, a single flower cluster with six open flowers was selected and tagged for monitoring.

The experiment followed a completely randomized design with five thrips density treatments: 0, 4, 8, 16, and 32 adult F. tritici per flower cluster. Each treatment was replicated ten times per variety, resulting in a total of 50 cages per variety. Thrips were collected from nearby flowering blueberry fields one day prior to infestation using a beating sheet and aspirator. Only adult female thrips were selected for use in the experiment. Prior to infestation, selected flowers were hand-pollinated to ensure fruit set. Thrips were then artificially introduced into treatment cages at the specified densities, while control cages received no thrips.

Data were collected at three developmental stages: flower stage, fruit set stage, and harvest stage. At the flower stage (72 hours post-infestation), the number of dropped flowers was recorded to calculate flower drop rate, and flower injury severity was rated on a 0–3 ordinal scale based on visible damage symptoms. At the fruit set stage, the number of successfully developed fruits was recorded per flower to determine the fruit set rate and bloom-to-fruit ratio. At harvest, fruits were evaluated for injury severity on a 0–3 scale and measured for weight (using a precision scale in grams) and diameter (using digital calipers in millimeters).