DISCRIMINATING BLACK AND WHITE SPRUCE POLLEN USING LAYERED MACHINE LEARNING

Pollen is among the most ubiquitous of terrestrial fossils, preserving an extended record of vegetation change. However, this temporal continuity comes with a taxonomic trade-off, with species-level identifications particularly elusive. We developed a supervised, layered, instance-based machine-learning classification system that uses leave-one-out bias optimization and discriminates among small variations in pollen shape, size and texture to address the problem of congeneric classifications. We tested our system on black and white spruce, two paleoclimatically significant species in the North American Quaternary. We achieved >93% grain-to-grain classification accuracies in a series of experiments with both fossil and reference material. More significantly, when applied to Quaternary samples, the learning system was able to replicate the count proportions of a human expert (R² = 0.78, p = 0.007), with one key difference – the machine achieved these ratios by including larger numbers of grains with low-confidence identifications. Ongoing research explicitly tests the ability of the learning system to capture known spruce pollen ratios against that of recognized experts.