Abstract Sedimentary pollen offers excellent opportunities to reconstruct vegetation changes over past millennia. Number of different pollen taxa or pollen richness is used to characterise past plant richness. To improve the interpretation of sedimentary pollen richness, it is essential to understand the relationship between pollen and plant richness in contemporary landscapes. This study presents a regional‐scale comparison of pollen and plant richness from northern Europe and evaluates the importance of environmental variables on pollen and plant richness. We use a pollen dataset of 511 lake‐surface pollen samples ranging through temperate, boreal and tundra biomes. To characterise plant diversity, we use a dataset formulated from the two largest plant atlases available in Europe. We compare pollen and plant richness estimates in different groups of taxa (wind‐pollinated vs. non‐wind‐pollinated, trees and shrubs vs. herbs and grasses) and test their relationships with climate and landscape variables. Pollen richness is significantly positively correlated with plant richness ( r = 0.53). The pollen plant richness correlation improves ( r = 0.63) when high pollen producers are downweighted prior to estimating richness minimising the influence of pollen production on the pollen richness estimate. This suggests that methods accommodating pollen‐production differences in richness estimates deserve further attention and should become more widely used in Quaternary pollen diversity studies. The highest correlations are found between pollen and plant richness of trees and shrubs ( r = 0.83) and of wind‐pollinated taxa ( r = 0.75) suggesting that these are the best measures of broad‐scale plant richness over several thousands of square kilometres. Mean annual temperature is the strongest predictor of both pollen and plant richness. Landscape openness is positively associated with pollen richness but not with plant richness. Pollen richness values from extremely open and/or cold areas where pollen production is low should be interpreted with caution because low local pollen production increases the proportion of extra‐regional pollen. Synthesis. Our results confirm that pollen data can provide insights into past plant richness changes in northern Europe, and with careful consideration of pollen‐production differences and spatial scale represented, pollen data make it possible to investigate vegetation diversity trends over long time‐scales and under changing climatic and habitat conditions.
Abstract Aim Pollen assemblages are commonly used to reconstruct past climates yet have not yet been used to reconstruct past human activities, including deforestation. We aim to assess (i) how pollen assemblages vary across biogeographic and environmental gradients, (ii) the source area of pollen assemblages from lake sediment samples and (iii) which pollen taxa can best be used to quantify deforested landscapes. Location Amazonia. Taxon Plantae. Methods Pollen assemblages ( N = 65) from mud‐water interface samples (representing modern conditions) of lake sediment cores were compared with modern gradients of temperature, precipitation and elevation. Pollen assemblages were also compared with local‐scale estimates of forest cover at 1, 2, 5, 10, 20 and 40 km buffers around each lake. Results Over 250 pollen types were identified in the samples, and pollen assemblages were able to accurately differentiate biogeographic regions across the basin, corresponding with gradients in temperature and precipitation. Poaceae percentages were the best predictor of deforestation, and had a significant negative relationship with forest cover estimates. These relationships were strongest for the 1 km buffer area, weakening as buffer sizes increased. Main conclusions The diverse Amazonian pollen assemblages strongly reflect environmental gradients, and percentages of Poaceae best reflect local‐scale variability in forest cover. Our results of modern pollen‐landscape relationships can be used to provide a foundation for quantitative reconstructions of climate and deforestation in Amazonia.
Predicting the trajectory of ongoing diversity loss requires knowledge of historical development of community assemblages. Long-term data from palaeoecological investigations combined with key biodiversity measures in ecology such as taxonomic richness, functional diversity (FD), phylogenetic diversity (PD) and environmental factors expressed as Ellenberg indicator values (EIVs) could provide that knowledge. We explored the modern pollen–plant (moss polster pollen vs surrounding vegetation) diversity relationships for herbaceous and woody taxa in calcareous fens from two different regions in Estonia, NE Europe. Associations of taxonomic richness, vegetation composition, FD (including functional alpha diversity and trait composition), PD and EIVs in modern pollen vs plant data were studied with correlation analysis, Procrustes analysis and linear regression models. To test their potential use in palaeoreconstructions, diversity measures were applied on pollen data from Kanna spring fen reflecting fen vegetation development over the last nine millennia and diversity changes through time were studied using generalized additive models. Results showed significant pollen–plant richness correlations for herbaceous taxa at vegetation estimate scales up to 6 m radius and Procrustes analysis showed significant compositional associations at all plant estimate scales (up to 100 m). Woody taxa had no significant pollen–plant richness correlations but composition relationships were significant at plant estimate scales of 6–100 m. Traits that were best reflected by pollen data (both in terms of trait composition and functional alpha diversity) among woody and herbaceous taxa were seed number, clonality, SLA and LDMC. PD of herbaceous species was reflected by pollen data. Among the EIVs, Ellenberg L and T were significantly reflected by pollen data for both woody and herbaceous communities. Palaeoreconstruction from Kanna fen indicates that trends of woody taxa are mostly related to long-term changes in climate while diversity variables of herbaceous taxa closely follow autogenic processes within the fen. We suggest that pollen-based diversity estimates should be calculated separately for woody and herbaceous taxa as they clearly represent different spatial scales. Present study suggests that linking sedimentary pollen data with FD, PD and EIVs provides possibilities to examine long-term trends in community assembly and ecosystem processes that would be undetectable from traditional pollen diagrams.
Abstract Human‐induced activities around Lake Lilaste in the central Latvian sandy coastal area have been reconstructed over a 1300‐year period. We use a combination of well‐established geoarchaeological research methods ( 14 C dating, pollen, nonpollen palynomorphs, REVEALS modeling, diatoms, C/N ratio, magnetic susceptibility, loss on ignition) to study the human impact on the environment. Historical context aids focus on records of resource (e.g., timber) exploitation in the area. The continuous record of human indicator pollen and agricultural landscape suggests this area was suitable for habitation well before the studied time period, likely due to the ecosystem services it provided. Our proxy‐based study, combined with historical background, reveals a significant human impact on the terrestrial environment since the 14th century. Deficiency of trees in the northern outskirts of Riga during the 17th–19th century was likely. Anthropogenic activity has led to both deforestation and change in species composition. Our paleo records indicate recognizable human‐driven legacy in current seacoast landscape.
