Variability within the 10-year seasonal Neotropical pollen rain and its implications for paleoenvironmental reconstructions

Background/Question/Methods

Compositional changes in Neotropical fossil pollen records are often interpreted using modern pollen rain samples collected from a range of different ecosystems. These changes are assumed to reflect community turnover, and paleoclimates and paleoenvironments are reconstructed based on taxonomic similarities between the modern and fossil assemblages. However, short-term, modern samples often do not capture the natural variability in pollen rain composition that is independent of changes in standing vegetation. Year-to-year variation in pollen influx is a result of both climatic variability and the endogenous phenological behavior of individual plants. To determine the extent to which short-term environmental changes affect pollen influx composition, we sampled ten years of seasonal pollen rain from Barro Colorado Island, Panama at a range of height through the canopy and compared the pollen influx to ten years of climatic data using canonical correspondence analysis. Time-averaged calculations of pollen influxes were compared using detrended correspondence analysis to estimate the sampling duration needed to capture the most representative pollen assemblage of the surrounding forest. “R-rel” – a measure of over- or underrepresentation of plant taxa in pollen assemblages – was used to assess how representative annual samples of the pollen rain were of the surrounding vegetation. 

Results/Conclusions

Annual calculations of “R-rel” demonstrated that annual pollen assemblages were disproportionately biased by the phenological behavior of individuals within 5 m of pollen traps; therefore, modern pollen sampling protocols that fail to account for distance to nearest reproductive individual will easily over- or underestimate the long-term R-rel value of individual plant taxa. Canonical correspondence analysis of annual pollen influxes revealed that even small annual differences in temperature, precipitation and light availability affected the composition and abundance structure of pollen samples. Annual variation in pollen composition was most affected by the preceding year’s precipitation, followed by the current year’s temperature and photosynthetically active radiation (PAR). Furthermore, comparisons of time-averaged pollen influxes within the ten years of pollen rain sampled suggest that a minimum of five years of continuously sampled pollen influx are needed to estimate the full range of natural variability in the pollen output to a single site for fossil record inferences.