Re-examination of the statistical methods used to determine the number of point counts needed for micropaleontological quantitative research

Abstract

Currently some controversy exists in the micropaleontological community concerning the statistically correct number of counts required for quantitative examinations, particularly with respect to the effect of variations in the number of species between samples and the significance of varying fractional abundances on the reliability of results. This analysis of the various statistical methods used to determine the number of required counts has shown that the number of species has no relationship to the number of counts required to measure accurately fractional abundances. As part of the study, logarithmic contours plotting percentage abundance against the total number of specimens, which provide abundance errors at a 95 percent confidence level, have been generated. The plot is displayed logarithmically to emphasize the significance of rare microfossil elements that dominate most assemblages, and which are important in many paleoenvironmental studies. Based on the plot, it is recommended that researchers utilize counts of at least 50 for indicator species having a fractional abundance of approximately 50 percent or greater; 300 counts for species which comprise approximately 10 percent of a sample; 500–1,000 counts for species that make up 5 percent of a sample; and counts of several thousand for defining species that comprise 1 percent of a sample. It is important to note, however, that where similar biofacies are involved, higher counts are required to accurately distinguish them. It is also recommended that researchers include fractional error abundances with their estimated abundances to provide an indication of their accuracy.