Martin Theuerkauf

The European Modern Pollen Database (EMPD) project

Modern pollen samples provide an invaluable research tool for helping to interpret the quaternary fossil pollen record, allowing investigation of the relationship between pollen as the proxy and the environmental parameters such as vegetation, land-use, and climate that the pollen proxy represents. The European Modern Pollen Database (EMPD) is a new initiative within the European Pollen Database (EPD) to establish a publicly accessible repository of modern (surface sample) pollen data. This new database will complement the EPD, which at present holds only fossil sedimentary pollen data. The EMPD is freely available online to the scientific community and currently has information on almost 5,000 pollen samples from throughout the Euro-Siberian and Mediterranean regions, contributed by over 40 individuals and research groups. Here we describe how the EMPD was constructed, the various tables and their fields, problems and errors, quality controls, and continuing efforts to improve the available data.

Pollen productivity estimates strongly depend on assumed pollen dispersal

Past plant abundance may be reconstructed from pollen data if dispersal distances of pollen and pollen productivities of each taxon are known. Using surface sediment samples from small and medium sized, closed and near circular lakes from lowland Central Europe, we tested the validity of three pollen dispersal models by comparing empirical pollen data from each lake with simulated pollen data derived from applying various pollen dispersal models to vegetation data from rings situated up to 100 km from each site. Pollen assemblages simulated with a Lagrangian stochastic (LS) model best fit real pollen assemblages, simulations with the commonly used Prentice model on pollen dispersal underestimated the amount of pollen arriving from distances larger than 10 km and overestimated the differences in dispersal distances between lighter (Pinus) and heavier (Fagus, Picea) pollen grains. The LS model appeared to provide more appropriate simulations. Pollen productivity estimates (PPEs) calculated for the data set showed that the choice of the dispersal model has great impact on the results. If derived with the Prentice model, PPEs for Fagus and Picea are three times higher than with the LS model. Studies on pollen productivities thus need to consider the apparent limitations of the Prentice model. We suggest an alternative approach, which uses simulations instead of the extended R-value model, to calculate PPEs. The approach is flexible in the use of dispersal functions and produced consistent results for two independent data sets from small and medium sized lakes.

Assessing resilience in long-term ecological data sets

In this paper the concept of resilience is discussed on the base of 13 case studies from the German branch of the International Long-Term Ecological Research Program. In the introduction the resilience approach is presented as one possibility to describe ecosystem dynamics. The relations with the concepts of adaptability and ecological integrity are discussed and the research questions are formulated. The focal research objectives are related to the conditions of resilient behaviour of ecosystems, the role of spatio-temporal scales, the differences between short- or long-term dynamics, the basic methodological requirements to exactly define resilience, the role of the reference state and indicators and the suitability of resilience as a management concept. The main part of the paper consists of 13 small case study descriptions, which demonstrate phase transitions and resilient dynamics of several terrestrial and aquatic ecosystems at different time scales. In the discussion, some problems arising from the interpretation of the time series are highlighted and discussed. The topics of discussion are the conceptual challenges of the resilience approach, methodological problems, the role of indicator selection, the complex interactions between different disturbances, the significance of time scales and a comparison of the case studies. The article ends with a conclusion which focuses on the demand to link resilience with adaptability, in order to support the long-term dynamics of ecosystem development.

A Lateglacial palaeosol cover in the Altdarss area, southern Baltic Sea coast (northeast Germany): investigations on pedology, geochronology and botany

A new site with Lateglacial palaeosols covered by 0.8 - 2.4 m thick aeolian sands is presented. The buried soils were subjected to multidisciplinary analyses (pedology, micromorphology, geochronology, dendrology, palynology, macrofossils). The buried soil cover comprises a catena from relatively dry (’Nano’-Podzol, Arenosol) via moist (Histic Gleysol, Gleysol) to wet conditions (Histosol). Dry soils are similar to the so-called Usselo soil, as described from sites in NW Europe and central Poland. The buried soil surface covers ca. 3.4 km 2 . Pollen analyses date this surface into the late Allerod. Due to a possible contamination by younger carbon, radiocarbon dates are too young. OSL dates indicate that the covering by aeolian sands most probably occurred during the Younger Dryas. Botanical analyses enables the reconstruction of a vegetation pattern typical for the late Allerod. Large wooden remains of pine and birch were recorded.

A matter of dispersal: REVEALSinR introduces state-of-the-art dispersal models to quantitative vegetation reconstruction

The REVEALS model is applied in quantitative vegetation reconstruction to translate pollen percentage data from large lakes and peatlands into regional vegetation composition. The model was first presented in 2007 and has gained increasing attention. It is a core element of the Landcover 6k initiative within the PAGES project. The REVEALS model has two critical components: the pollen dispersal model and pollen productivity estimates (PPEs). To study the consequences of model settings, we implemented REVEALS in R. We use a state-of-the-art Lagrangian stochastic dispersal model (LSM) and compare model outcomes with calculations based on a conventional Gaussian plume dispersal model (GPM). In the LSM turbulence causes pollen fall speed to have little effect on the dispersal pattern whereas fall speed is a major factor in the GPM. Dispersal models are also used to derive PPEs. The unrealistic GPM produces PPEs that do not describe actual pollen productivity, but rather serve as a basin specific correction factor. A test with pollen and vegetation data from NE Germany shows that REVEALS performs best when applied with the LSM. REVEALS applications with the GPM can produce realistic results, but only if unrealistic PPEs are used. We discuss the derivation of PPEs and further REVEALS applications. Our REVEALS implementation is freely available as the ‘REVEALSinR’ function within the R package DISQOVER. REVEALSinR offers an environment for experimentation and analysing model sensitivities. We encourage further experiments and welcome comments on our tool.

Varve microfacies and varve preservation record of climate change and human impact for the last 6000 years at Lake Tiefer See (NE Germany)

The Holocene sediment record of Lake Tiefer See exhibits striking alternations between well-varved and non-varved intervals. Here, we present a high-resolution multi-proxy record for the past ~6000 years and discuss possible causes for the observed sediment variability. This approach comprises microfacies, geochemical and microfossil analyses and a multiple dating concept including varve counting, tephrochronology and radiocarbon dating. Four periods of predominantly well-varved sediment were identified at 6000–3950, 3100–2850 and 2100–750 cal. a BP and AD 1924–present. Except of sub-recent varve formation, these periods are considered to reflect reduced lake circulation and consequently, stronger anoxic bottom water conditions. In contrast, intercalated intervals of poor varve preservation or even extensively mixed non-varved sediments indicate strengthened lake circulation. Sub-recent varve formation since AD 1924 is, in addition to natural forcing, influenced by enhanced lake productivity due to modern anthropogenic eutrophication. The general increase in periods of intensified lake circulation in Lake Tiefer See since ~4000 cal. a BP presumably is caused by gradual changes in the northern hemisphere orbital forcing, leading to cooler and windier conditions in Central Europe. Superimposed decadal- to centennial-scale variability of the lake circulation regime is likely the result of additional human-induced changes of the catchment vegetation. The coincidence of major non-varved periods at Lake Tiefer See and intervals of bioturbated sediments in the Baltic Sea implies a broader regional significance of our findings.

Effects of changes in land management practices on pollen productivity of open vegetation during the last century derived from varved lake sediments

Pollen productivity is a key parameter to quantify past plant abundances and vegetation openness. In this study we explore how changes in land management influence pollen productivity. We study pollen deposition in largely annually laminated sediments from Lake Tiefer See in the Northeastern German lowlands deposited between AD 1880 and 2010. During this period, land use intensity has increased predominantly through the widespread introduction of artificial fertilizers, herbicides and heavy machinery mainly since the 1950s. Although land use statistics show that overall vegetation openness remained largely constant, non-arboreal pollen deposition (from herbs and grasses) sharply declined over the study period. This decline can be partly explained by a shift towards crops that emit little pollen such as wheat and oilseed rape. Furthermore, intensified grassland management, including drainage, also contributed to lower pollen deposition because of the decline of Plantago lanceolata and Rumex from grassland communities. However, the most important effect is a decline in pollen productivity of grasses of about 60%, which most likely is a response to earlier and more frequent mowing, although changes in grass species composition may also have played a role. Our results show that the type and intensity of land use have a strong effect on pollen productivity of grasses (and smaller effects on further crops). Since grass pollen deposition is a main proxy for vegetation openness and grasses are the common reference taxon in most PPE studies, variations in the pollen productivity of grasses introduce so far neglected errors in the reconstruction of the past vegetation cover. Our study provides a first estimate about the magnitude of this effect.

Understanding the Long Term Ecosystem Stability of a Fen Mire by Analyzing Subsurface Geology, Eco-Hydrology and Nutrient Stoichiometry – Case Study of the Rospuda Valley (NE Poland)

We explored the background of differences in long–term stability between two parts in an undisturbed mire system (Rospuda fen, NE Poland). We re-constructed the Holocene history of the mire and compared it with current vegetation, water level dynamics, water chemistry and nutrient availability in two basins: A, where the mire terrestrialised a deep gyttja–filled lake, and B, where peatland developed directly on fluvial sands. The current vegetation of sedge–moss fens was described in 10 relevés from each basin, groundwater was sampled from piezometers and analysed for major ions, while its relative water level was recorded during three years. N and P content was measured in above ground vascular plant samples collected within the relevés. Fens in basin A were stable in the past, whereas fens in basin B switched between open and wooded or reed–dominated phases. In basin B, where trees are more abundant, we found higher water fluctuations, occurrence of river floods and a higher N:P ratio than in basin A. Our interpretation follows that the subsurface geology of fen basin may govern mire stability by determining its hydrological–buffering capacity, which may affect N:P ratios. Our results suggest that P–limited fens are more vulnerable for changes in water level.

MARCO POLO – A new and simple tool for pollen-based stand-scale vegetation reconstruction

Hitherto, the Landscape Reconstruction Algorithm (LRA) has been the only truly quantitative approach to stand-scale palynology. However, the LRA requires information on pollen productivity and dispersal, which is not always available. The alternative approach MARCO POLO (MAnipulating pollen sums to ReCOnstruct POllen of Local Origin) presented here is solely based on pollen values and does not rely on a pollen dispersal function. In a stepwise fashion, MARCO POLO removes those pollen types from the pollen sum whose values are significantly higher than in a neighbouring large basin. The resulting regional pollen sum is free of the disturbing factor of (extra-)local pollen. Based on this sum, comparison with the pollen record from the large basin allows calculating sharp (extra-)local signals. Treating the (extra-)local pollen portion with representation factors (R-values) then produces a quantitative reconstruction of the stand-scale vegetation composition. We tested MARCO POLO and the LRA on a dataset of pollen surface samples and forest vegetation relevés from northern Central Europe. Both approaches reconstruct the presence or absence of taxa at the stand scale within a small margin of error. Where observed cover was ⩾2%, both models always reconstructed presence, where modelled cover was ⩾2% the taxon was always present. Overall, both approaches perform well in reconstructing the cover of taxa within a 100-m radius. In our tests, MARCO POLO is slightly better at reconstructing cover values for more taxa. Although some model parameters evidently need revision, the simple correlative approach of MARCO POLO appears to perform at least as well as the complex LRA model.

The extended downscaling approach: A new R-tool for pollen-based reconstruction of vegetation patterns

The extended downscaling approach (EDA) is a quantitative method in palynology that aims to detect past vegetation patterns and communities in the landscape. The EDA uses iterative forward modelling to fit vegetation composition to robust landscape patterns by comparing simulated with actually observed pollen deposition. The approach employs a set of pollen records, preferably from medium sized to large lakes or peatlands, as well as maps of robust landscape patterns, such as soils and relief. So far, the EDA has been applied in simple settings with only few taxa. To be able to apply the model also in more complex situations, we have implemented the EDA in the R environment for statistical computing. We here test the performance of the EDAinR function in five synthetic scenarios of increasing complexity. In all cases, the EDA is well able to reconstruct vegetation composition, also on rare landscape units. If uncertainty is added both to the pollen data and pollen productivity estimates, the EDA still correctly reconstructs species composition on more than 90% of the total landscape in all scenarios, underlining that the EDA performs well also in complex settings. The EDAinR function will be available within the R package DISQOVER.