Characteristics Of Different Plant Communities And Their Relationships With Surface Pollen Assemblages In The Lhasa Valley

In order to study the structural characteristics of different plant communities and their relationships with surface pollen assemblages in the Lhasa Valley, 47 sampling plots were investigated, three vegetation quadrats and at least five surface pollen samples of each plot were collected during field work, five surface pollen samples were mixed into one sample afterwards, namely our study was based on a dataset of 141 vegetation quadrats and 47 pollen samples.Based on climatic information of each plot derived from 16 meteorological stations in and surround the Lhasa Valley using Co-Kriging methods, mean annual precipitation(Pann), mean annual temperature(Tann), summer temperature(Tsummer), actual evapotranspiration(AET), potential evapotranspiration(PET) and moisture index(MI) were calculated. Using Cluster analysis to classify different plant communities, DCA(Detrended Correspondence Analysis) to examine the relationships between plant communities and its relation to environmental factors, GAM(Generalized Additive Model) and Correlation analysis to understand the relationships of TSR(Transformed species richness) and environmental factors, Cluster analysis and PCA(Principal Components Analysis) to analyze the relationship between surface pollen assemblages and vegetation, Univariate linear regression analysis to examine relationships between pollen ratios(A/C, aridity pollen ratio, AP, A/Cy, AC/Cy) and climatic factors.Cluster analysis of plant communities showed 8 vegetation groups, including a group of scrub-meadow(SM) and alpine meadow(M), two groups of shrubs(S1, S2), coniferous wood(CW), deciduous broad-leaf wood(DBW) and shrub-grassland(SG). Result of DCA ordination was similar with the result of cluster analysis, except in DCA ordination, alpine desert(AD) was identifiable from others and shrubs(S) was scattered with no tendency of grouping. Correlation of DCA axes and environmental factors indicates hydrothermal factors are the most important factors that influent the structural characteristics of plant communities and their distribution patterns in the Lhasa valley. First is water conditions, and the second important factors are thermal factors, disturbance could also be one of the important factors in the study area.The results of cluster analysis and PCA of pollen samples are generally consistent, except for PCA could differentiate shrubs slightly better. Both of them could well distinguish vegetation of arid conditions and vegetation from relatively humid conditions, and there is a better agreement between surface pollen samples and the distribution of deciduous broadleaf wood than of coniferous wood. PCA of both pollen assemblages and vegetation shows that the first axis of each PCA results reflects the same humidity gradient, namely humidity is the main influencing factor to both modern pollen assemblages and vegetation in Lhasa Valley. Thermal factor is another important environmental factor to pollen assemblages.A/C ratio and aridity pollen index both could not be reliable aridity indicators in Lhasa Valley. Though AP sum could be a precipitation indicator, however it is less reliable than A/Cy ratio and AC/Cy ratio. A/Cy ratio and AC/Cy ratio could be more reliable precipitation, aridity and temperature indicators in the Lhasa Valley.