Search

Past plant diversity dynamics can provide a historical perspective on biodiversity change and ecosystem dynamics as a guide to conservation management in the future. Pollen analysis is one of the most widely used approaches to reconstruct plant diversity change through time, but the representativeness of pollen diversity to plant diversity is not yet fully understood and especially in open landscapes. In this paper, we compare pollen assemblages from topsoil in the open landscapes of the Tibetan Plateau with the surrounding vegetation to investigate their compositional similarity and the correlations between their diversity (richness and evenness), and to evaluate the influence of climate variables, landscape characteristics, and human disturbance on the pollen diversity. The results show that pollen assemblages vary obviously in different vegetation types and they can represent their surrounding plant communities at low taxonomic levels. Pollen richness boarders a better match with plant richness than pollen evenness with plant evenness across the whole study area, but the correlation is weak (r adj 2 < 0.2, p < 0.001). Transforming plants into their pollen equivalents has little improvement for the correlations between pollen and plant richness/evenness. Pollen richness and evenness can capture well the trend in plant richness and evenness along the vegetation gradient, suggesting that they can be used as gamma (landscape scale) rather than alpha (local scale) diversity metrics in the study area. Mean annual precipitation (MAP), mean annual temperature (MAT), and patch richness (PatchS) are positively associated with pollen richness. MAP, PatchS, and human disturbance (HD) are significantly predictors for pollen evenness. These confirm that pollen richness and evenness can provide insights into past plant diversity dynamics and help evaluating the effects of climate and habitat changes and human disturbance on long timescales. • Pollen assemblages represent surrounding plant communities at low taxonomic levels. • Pollen diversity is reliable proxy for plant diversity at the landscape scale. • Vegetation shift is an important prerequisite in studying past plant diversity. • Pollen-equivalent transformation cannot improve pollen-plant diversity correlations. • Pollen diversity can indicate changes in past climate and habitat conditions. [ABSTRACT FROM AUTHOR]

Understanding human impact on the vegetation of the Tibetan Plateau through the Holocene is important for developing strategies for sustainable ecosystem management in the region. In this paper, 53 pairs of surface-soil samples from inside and outside livestock enclosures were obtained from across the Tibetan Plateau to investigate the differences in pollen assemblages between grazed and ungrazed sites. This modern dataset was then applied to palynological records from four lakes in the northeast Tibetan Plateau region to investigate middle to late Holocene grazing history. Results show that modern grazing activities have a limited impact on the vegetation, although grazing activity leads to an increase in the pollen proportion of Cyperaceae, Artemisia , Ranunculaceae, Brassicaceae, Taraxacum -type, Fabaceae, and Saxifrageceae while representatives of Poaceae, Chenopodiaceae, and Aster -type pollen decrease. Application of these grazing indicators to the pollen spectra obtained from the four lake sites, together with other proxies, reveals that grazing activity only commenced during the last few hundred years in the northeast marginal areas of the Tibetan Plateau at c. 3000 m a.s.l, while there is no grazing signal in pollen records from above 4000 m a.s.l. Our study emphasises the importance of using multi-proxy methods in estimating signals of past human activity. • Pollen assemblage should be insensitive to grazing activities on the Tibetan Plateau. • Human impacts on vegetation at ∼3400 m a.s.l are limited to within a few centuries. • No significant pollen signal of grazing is revealed for the high-elevation areas. [ABSTRACT FROM AUTHOR]