Relevant source area of pollen and relative pollen productivity estimates in Bashang steppe

Relative pollen productivity estimates (RPPEs) are indispensable prerequisites for quantitative vegetation reconstructions. RPPEs have now been calculated in many regions of Europe, while only few work has been done in China. In this paper, the relevant source area of pollen (RSAP) and the RPPEs of the common taxa of Bashang steppe (41°10'~42°50'N, 114°10'~116°10'E) were done with 3 Extended R-Value (ERV) models using ERV-Analysis .v1.2.3 Program. Pollen sampling was performed at 30 random sampling sites selected by random function. Vegetation survey followed the standard vegetation survey method promoted by "Crackles Bequest Project". Top soils are used for the pollen samples. The pollen analysis results showed that Artemisia, Chenopodiaceae, Poaceae, Asteraceae and Cyperaceae etc. herbaceous taxa dominate pollen assemblages. That is basically consistent with composition of the steppe vegetation. However, the differences are obvious between the pollen percentage and vegetation cover. So we calculate the RASP and RPPEs to calibrate the relationship between pollen percentage and vegetation cover. The RSAP is about 2100m in Bashang steppe based on ERV submodel 3 with Prentice's distance-weighting method. The comparison of RASP in different wind speeds indicates that the RSAP changes very few with the increasing of wind speed from 3.5m/s. The results of RPPEs suggest that the values of Artemisia and Chenopodiaceae are 20 times higher than Poaceae, Cyperaceae and Asteraceae are 7~9 times, Liliaceae is only slightly higher than Poaceae, Convolvulaceae, Fabaceae, Lamiaceae and Potentilla are much lower than Poaceae. Then the relationship between pollen and vegetation are calibrated with RPPEs. The results show that the relationship between pollen and vegetation will become much better after calibrated with RPPEs for Artemisia and Chenopodiaceae because their higher percentages appear both in vegetation and pollen assemblages, thus their RPPEs can be used as a reference of quantitative reconstruction of paleovegetation, while the calibration results with RPPEs for Cyperaceae, Asteraceae, Liliaceae and Lamiaceae are not very good because their higher percentage only exist in vegetation or pollen assemblages, thus more work should be done if we want to use their RPPEs in quantitative reconstruction of paleovegetation. On account of low percent at the pollen assemblages (lower than 5 %), the calibration results are not obvious for other pollen taxa.