Results of 3 years of pollen trapping samples close to large canopy gaps had exceptional abundances of 'disturbance' taxa. A preliminary on Barro Colorado Island, Panama, are presented. Minimum pollen dispersal distances are estimated for analysis of the representation of canopy components indicates that as much as 19% of pollen caught in the the most abundant pollen taxa. Dispersal distances for some taxa appear to be as low as 5m, while for other traps was derived from large tree species. taxa at least 50% of the trapped pollen had travelled more than 40m. No consistent pattern of spatial Key words. Barro Colorado Island, Panama, pollen dispersal, pollen trap, tropical rain forest. variation was reflected in the pollen rain, however, 0-5m Annona,

Three years of pollen trapping data from Barro Colorado Island, Panama, are compared with local vegetation inventories. Two hypotheses relating pollen representation to 'messy' pollination and flower form are tested. Dioecious taxa were found to be over-represented in pollen spectra compared with their occurrence in local forests. Similarly, anemophilous and 'messy' pollination types were found to be over-represented. While anemophilous taxa were the best dispersed pollen types, zoophilous taxa were also well-represented in dispersal classes of 20-40 m and > 40 m. Thus pollen arriving to lake sediments is most likely to be from anemophilous species or those zoophilous species exhibiting 'messy' pollination syndromes. Pollination mechanisms will predictably bias the fossil record against certain flower morphologies. These data are of significance to those using the fossil pollen record to reconstruct the timing and sequence of angiosperm evolution. These data help prioritize plants to be included in modern pollen reference collections and to focus the search for 'unknown' types on most-likely candidate families.

Results of 3 years of pollen trapping on Barro Colorado Island, Panama, are presented. Minimum pollen dispersal distances are estimated for the most abundant pollen taxa. Dispersal distances for some taxa appear to be as low as 5 m, while for other taxa at least 50% of the trapped pollen had travelled more than 40 m. No consistent pattern of spatial variation was reflected in the pollen rain, however, samples close to large canopy gaps had exceptional abundances of ‘disturbance’ taxa. A preliminary analysis of the representation of canopy components indicates that as much as 19% of pollen caught in the traps was derived from large tree species.

No abstract