Assessing the influence of vegetation structure and phenological variability on pollen‐vegetation relationships using a 15‐year Neotropical pollen rain record

• Published in: Journal of vegetation science 2020-07, Vol.31 (4), p.606-615
• Main Authors: Haselhorst, Derek S., Moreno, J. Enrique, Punyasena, Surangi W., Giesecke, Thomas
• Format: Article
• Language: English
• Subjects: Annual rainfall; BCI; Fossil pollen; Lake sediments; Neotropical; phenology; Plant communities; Pollen; Questions; Rain; relationships; R‐rel; Sampling; Sediment samplers; Traps; tropical; Variability; Vegetation; Windows (intervals)
• ISSN: 1100-9233; 1654-1103
• DOI: 10.1111/jvs.12897

Questions Pollen collected from aerial pollen traps are used to interpret compositional changes in the fossil pollen record. Relationships between pollen abundances and vegetation are commonly measured using the ratio between the abundance of a pollen type and its corresponding basal area in a plant community (R‐rel). Pollen–vegetation relationships have been extensively studied in temperate and boreal latitudes using surface sediment, moss polster and pollen trap samples, representing long‐term accumulations of pollen. In contrast, pollen–vegetation relationships are not well‐resolved in diverse tropical habitats because of a lack of modern long‐term pollen and vegetation datasets. In the present study, we address two primary questions: 1. How variable are tropical pollen–vegetation relationships measured using R‐rel over time? 2. To what extent are differences in pollen abundances among traps related to variability in basal area in the surrounding community? Location Barro Colorado Island, Panama. Methods Variability in an annually sampled, 15‐year pollen rain was compared with annual basal area data collected over the same timeframe using R‐rel. Results R‐rel values were highly variable on a year‐to‐year basis and more consistent over time‐averaged sampling windows ≥3 years. A strong positive relationship between pollen abundances and local basal area surrounding pollen traps was determined in eight taxa. Conclusions The pollen–vegetation relationships provide a summary of the pollen representation of Neotropical taxa common in paleoecological samples over an extended sampling duration that approaches the timespan represented in lake sediment samples (~20 years). The results highlight taxa in which differences in pollen abundances among traps were strongly correlated with changes in surrounding basal area. This study presents a long‐term (1994–2009) assessment of pollen–vegetation relationships in a Neotropical forest using the ratio R‐rel (% relative pollen abundance/% relative basal area). Using annual pollen abundances obtained from 10 pollen traps and concurrent vegetation records, we evaluated how pollen–vegetation relationships for different taxa are influenced by phenological variability in pollen abundances and spatial variability in basal area.