Evaluation of differential pollen deposition and pollen focussing from three Holocene intervals in sediments from Lake O'Hara, Yoho National Park, British Columbia, Canada: intra-lake variability in pollen percentages, concentrations and influx

Abstract

Pollen spectra from three discrete intervals, dated by tephras and radiocarbon, are examined in a lake-wide series of nineteen cores from Lake O'Hara, Yoho National Park. These represent three different pollen assemblage zones recognized in the record of a single core ( LOH25). The three samples sets provide lake-wide pollen assemblages at ca. 2350, ca. 6800, and ca. 10,100 yr BP. These spectra are used to investigate two aspects of pollen accumulation: (1) whether pollen taxa have been differentially deposited in sediments in different parts of the lake basin, and (2) whether the entire assemblage has accumulated to a greater extent in deeper parts of the basin (“pollen focussing”). Percentages of the ten major taxa in the assemblages show remarkable basin-wide consistency within each sample set, suggesting that in this case a single core provides a representative pollen record. Multivariate statistical analyses (discriminant function analysis and analysis of variance) on pollen percentage, concentration, and influx data confirm that sample sets are distinctly different and support zonation of the pollen record from a single core. Minor components of the assemblages (< 1% of the pollen sum), mainly NAP, show great variability within and between sample sets and are therefore not reliable for interpretation of vegetation history. Lack of evidence for differential pollen deposition confirms that qualitative inferences based on major components of the pollen sum are probably reliable despite core location within the lake basin. The data do not display consistent evidence for greater pollen accumulation at deeper water sites implying that pollen focussing has not been a major process. Sedimentation rates are variable across the basin at each level, suggesting that pollen concentration and particularly pollen influx data may be unreliable for interpretation of vegetation history from a single core, without much greater chronologic control.