POLLEN ACCUMULATION IN A MONTANE REGION OF COLORADO, USA - A COMPARISON OF MOSS POLSTERS, ATMOSPHERIC TRAPS, AND NATURAL BASINS

Pollen spectra from three types of modern collectors - Tauber traps, moss polsters, and surface muds from small basins - are compared to illustrate pollen dispersal and sedimentation in an alpine setting. Pollen frequency data demonstrate that the atmospheric collectors and moss polsters provide the most distinct vegetation signals. In forest stands, pollen frequencies for each of the five main forest trees - Picea engelmannii, Abies lasiocarpa, Pseudotsuga menzeisii, Pinus contorta, and Populus tremuloides - are highest where the trees grow. Thus, each forest type can be recognized by its dominant arboreal taxa. The distinctions between vegetation types are less pronounced in pollen spectra from surface muds collected from small lakes. Surface lake samples collected above treeline have high frequencies of conifer pollen, particularly Pinus pollen. Picea and Abies pollen frequencies (averaging 16% and 2%, respectively) are somewhat lower above treeline than in the sub-alpine forest (where they average 20% and 7%). This suggests that the upper forest border can be recognized by its pollen spectrum. However, Pinus pollen above treeline is comparable to samples collected in the Pinus contorta forest (approximately 40%). The pollen accumulation rates in small bogs in forested settings and above treeline are comparable to the atmospheric pollen influx in these environments. The pollen influx in lakes surrounded by closed forests is only 10-20% greater than the accumulation rates in atmospheric collectors. Thus, for bogs in forests and krummholz, and for lakes without inflowing streams in forest stands, fluvial introduction of pollen is minimal. Atmospheric collectors and moss polsters may provide accurate modern analogues for fossil sites in these environments. However, atmospheric pollen accumulation accounts for only one-fifth of the pollen influx in an alpine lake without an inflowing stream, suggesting that in alpine tundra pollen is concentrated by slope wash across open steep terrain before it is deposited in small ponds. Because of this increased concentration, and because much of the pollen in alpine settings comes from plants growing at much lower elevations, alpine pollen spectra become badly distorted. Therefore, palynological inference of alpine tundra from fossil pollen spectra may be difficult. Discriminant analyses are used to test the degree of similarity between vegetation types according to their pollen spectra. While pollen spectra from moss polsters differ significantly by vegetation type, this is not true of pollen spectra from surface lacustrine samples. Bogs offer depositional settings comparable to moss polsters and atmospheric traps. All three of these can provide useful modern analogues for fossil pollen from bogs. In contrast, pollen spectra from small lakes, which tend to be less distinctive, particularly in open vegetation, may be difficult to interpret. Future calibration of modern pollen and vegetation with fossil pollen spectra should focus on bogs and small closed basin lakes without inflowing streams.