Analysis of spatial pattern of Frankliniella occidentalis (Thysanoptera: Thripidae) on greenhouse cucumbers using dispersion index and spatial autocorrelation

Studies were conducted in two commercial cucumber greenhouses to examine thrips' seasonal and spatial patterns on Cheju Island, Korea, in 1996. Leaf and flower samples were taken to determine the thrips species and stage complexes inhabiting cucumber plants. Adult thrips (54-55%) were the most dominant stage on flowers and the majority of adult species was Frankliniella occidentalis (Pergande), whereas immature (82-84%) was the dominant stage on leaves. To determine the spatial distribution patterns of thrips, at least 63 leaf samples which were regularly spaced within a greenhouse were visually inspected from among the Ist, the 7th and the 15th leaves assigned from the top plant canopy. The leaf positions were located at 1.8, 1.0 and 0.3 m above ground level, respectively. Taylor's power law indicated that the counts of thrips on leaves were aggregated regardless of thrips stage and leaf position. Autocorrelation analysis, which is based on the relative position of samples revealed the different spatial distribution patterns among the leaf positions. In general, the counts of thrips were nonrandomly distributed on the 7th leaf position, whereas the counts were randomly distributed on the ist and 15th leaf positions. Correlograms suggested the presence of single or multiple gradients within the sample portion of the greenhouse, depending on thrips stage. Our results suggest that Taylor's power law cannot detect the spatial relationship in data sets accurately, and testing correlograms for significance is more accurate for describing the spatial distribution patterns of thrips.