Sampling in aerobiology. Differences between traverses along the length of the slide in Hirst sporetraps

Abstract

Two years of data from four longitudinal traverses along each day's slide prepared from a continuously running Burkard sporetrap have been analyzed statistically. Using the Friedman test, a statistically significant difference was found between the four traverses, with a greater than 7% loss of pollen grains in the two outer traverses in relation to the inner. Four slides were then selected for more detailed analysis, using 18 longitudinal traverses with a 1-mm separation from the upper to the lower edge of the Melinex tape. There was found to be a progressive decline from the centre to the outside, and more than 4% of pollen grains were found outside the typical 14 mm width of the impaction orifice. There was no correlation between pollen grain size and the decline in counts from the centre to the outside. For the complete data set, there was a general rise in the diversity of bollen types with increasing sample counts, but above about 1000 pollen grains per sample there were no more than 27 pollen types found, often even fewer. A discussion is presented of whether four traverses really should be a fixing sample size, or whether it might be better to fix the total pollen count beginning with a traverse in the middle of the slide and ending with a variable number of traverses when that count is reached.