Aim: To provide a database of modern pollen abundances from a diversity of Neotropical locations.

Location: The Neotropics, especially western Amazonia, the Andes, The Galápagos, Central America, Puerto Rico, and Mexico.

Methods: Over a 30-year period, mud-water interface samples, moss polsters, soils, and pollen traps were used to assay modern pollen rain. Standard extraction methods were used and a minimum of 300 terrestrial pollen grains counted and identified per sample. Bioclimatic data were assigned to each location, and sexual traits and pollination syndromes are provided for some genera. A detrended correspondence analysis (DCA) shows the structuring of the data relative to mean annual temperatures (MAT), mean annual precipitation (MAP), and precipitation of the driest quarter (PDQ). Probability density functions of distances among samples are calculated within and between Holdridge Life Zones, based on sample DCA scores.

Results: The modern pollen of 636 locations are documented, with >500 pollen types identified. The dataset spans settings with MAT < 5oC to >25oC, and with MAP ranging from c. 500-5000 mm, and PDQ from 80 - 1500 mm. Ordination of the dataset provides evidence that assemblages differ according to temperature and precipitation, although historical biogeography also shapes pollen rains. The pollen grains of zoophilous species contribute important information to habitat descriptions, but they are under-represented in the dataset.

Main conclusions: The dataset covers a broad environmental range and can help inform biogeographic and paleoenvironmental interpretation of paleoecological records. Some caution in using the data is needed as the field of tropical palynology is maturing and the data gathered 30 years ago are less detailed than those gathered recently. Although anemophilous taxa tend to be over-represented in neotropical pollen spectra, the majority of pollen identified in each sample are from zoophilous species.

Modern pollen collected over ~30 years from mudwater interface, moss polsters, pollen traps and river mud. Standard palynological processing of samples prior to counting.

none