Pollen counts statistics and its relevance to precision
Tóm tắt
In the day to day management of pollen counts from aerobiological samples of national networks, only a small proportion (usually from 12 to 15%) of the daily microscope slide is read. It is generally believed that, otherwise, too much time will be spent reading slides for a minimal increase in precision. Different networks use different slide sampling methods (longitudinal, transverse or at random) and a different number of lines are routinely read. However, the topic of the precision of the different pollen count strategies has seldom been the object of serious investigation. In this study, the precision of different sampling methods of 12 pollen types was investigated by: a) counting pollen grains over the whole slide (3 slides per taxa), b) spatially (i.e. microscope field per microscope field) recording over the 3120 fields found at 400× the location of each pollen grain, c) sub-sampling, by macro procedures, this population by selecting a number of transverse (1 to 48) or longitudinal (1 to 20) lines, or a number of random fields (90 to 2340), so that between 0.96 to 46.15% (transverse), 3 to 66.6% (longitudinal) or 3 to 75% (random) of the whole slide was artificially counted. Between nine and twelve procedures were built per reading strategy. The error found is much higher than what is normally believed, and it was significantly correlated with the abundance of a pollen taxa on the sampled slide. It is only with a total count over 1000 (corresponding to a concentration of above 500 m−3) that the mean error of 4 longitudinal lines (or 13.3% of the slide), the standard protocol of both the Italian Association of Aerobiology (AIA) and the Spanish Aerobiology Network (REA), was always below 30%.
Tài liệu tham khảo
Barkeley F.: 1934, The statistical theory of pollen analysis. Ecology 15, 283–289.
Hill T.R.: 1996, Statistical determination of sample size and contemporary pollen counts, Natal Drakensberg, South Africa. Grana 35, 119–124.
Kapyla and Penttinen: 1981, An evaluation of the microscopic counting methods of the tape in Hirst-Burkard pollen and spore trap. Grana 20, 131–141.
Miquel P.: 1883, Les Organismes Vivants de l'Atmosphère. Gauthier-Villars, Paris, 310 pp.
Pedersen B.V. and Moseholm L.: 1993, Precision of the daily pollen count. Identifying sources of variation using variance component models. Aerobiologia 9, 15–26.
Regal R.R. and Cushing E.J.: 1979, Confidence intervals for absolute pollen counts. Biometrics 35(3), 557–565.
Rull V.: 1987, A note on pollen counting in palaecology. Pollen et Spores 24(4), 471–480.
Tormo Molina R., Munoz Rodriguez A. and Silva Palacios I.: 1996, Sampling in aerobiology. Differences between traverses along the lenght of the slide in Hirst spore trap. Aerobiologia 12, 161–166.