Allelic richness predicted by subsampling and regression modeling for microsatellite data. PR: Picea rubens, TO: Thuja occidentalis, PG: Picea glauca, PS: Pinus strobus. The population names are provided in Table 1. PG1-Subsampling - TO3-Subsampling: allelic richness estimated by repeated random subsampling from the amplified empirical data. PG1-Regression - TO3-Regression: regression curves for allelic richness predicted by equation (5). Regression curves correspond well with the allelic richness estimates obtained from subsampling of the actual data. Subsampling was performed by replicating the empirical data set up to n = 10,000, and randomly drawing samples of a given n from the amplified population, in 50 replicates.