Research on pollen productivity estimates as baseline information for quantitative reconstruction of vegetation during the Last Glacial Maximum

During the Last Glacial Maximum (LGM), pinaceous species such as those in Abies, Tsuga, Picea and Pinus were dominant in Japan (Tsukada, 1983). Although the quality and quantity of plaeoecological data for the LGM has increased throughout Japan, quantitative reconstruction of the LGM vegetation is still in its infancy because of the lack of pollen productivity estimates (PPEs) for those taxa – one of the essential parameters necessary for the Landscape Reconstruction Algorithm (LRA; Sugita, 2007a, b). This study uses the litter-trap method of Saito (Saito and Takeoka,1983, 1985) to obtain PPEs for thirteen species in Pinaceae and Betula characteristic for the LGM vegetation. We use 5-12 traps 50 cm × 50 cm in opening that are placed in each of nearly monospecific stands of those species in various parts of Japan. The number of male flowers produced per ha per year (Fn) is estimated from those collected in traps, and the number of pollen grains in male flowers (PFn) from flower specimens collected just before flowering and fixed in Farmer’s fixative (Ethanol : Acetic acid =3:1) for pollen counting later. PPEs [grains ha^-1 year^-1] are calculated as Fn･PFn. Also, evaluation of the LRA is currently underway in the subalpine forests of the Kita Yatsugatake mountains, using pollen data from the Shirakomaike pond (ca. 7.7 ha) and the Shirakoma mire (ca. 0.1 ha) and the vegetation composition around the sites. Year-to-year variations of pollen production for those species will be measured several more years to increase the reliability of the LRA-based reconstruction.