View article

The most popular method for reconstructing past environments is the method of modern analogs. As with any quantitative paleoenvironmental methodology, this technique requires a wide array of modern calibration samples spanning the entire range of present-day and paleo-environments. One advantage of the method is the wide range of analytical tools available for analysis of specific outcomes. Using as example North American pollen data, we show how dissimilarity analysis can be improved using mapping and multivariate analysis. Examination of the dissimilarity measure in temporal, geographic and taxonomic space provides reasons for choices of a particular modern analog for any fossil assemblage. This analysis quantifies extrapolations (non-analogs) and further provides methods for determining the causes.