Catalog of fossil spores and pollen,

Cumulative index: v. 11-20, 1965. v. 21-30, 1970. v. 31-40, 1977.

Impact of model complexity on cross-temporal transferability in Maxent species distribution models: An assessment using paleobotanical data

Maximum entropy modeling (Maxent) is a widely used algorithm for predicting species distributions across space and time. Properly assessing the uncertainty in such predictions is non-trivial and requires validation with independent datasets. Notably, model complexity (number of model parameters) remains a major concern in relation to overfitting and, hence, transferability of Maxent models. An emerging approach is to validate the cross-temporal transferability of model predictions using paleoecological data. In this study, we assess the effect of model complexity on the performance of Maxent projections across time using two European plant species (Alnus giutinosa (L.) Gaertn. and Corylus avellana L) with an extensive late Quaternary fossil record in Spain as a study case. We fit 110 models with different levels of complexity under present time and tested model performance using AUC (area under the receiver operating characteristic curve) and AlCc (corrected Akaike Information Criterion) through the standard procedure of randomly partitioning current occurrence data. We then compared these results to an independent validation by projecting the models to mid-Holocene (6000 years before present) climatic conditions in Spain to assess their ability to predict fossil pollen presence-absence and abundance. We find that calibrating Maxent models with default settings result in the generation of overly complex models. While model performance increased with model complexity when predicting current distributions, it was higher with intermediate complexity when predicting mid-Holocene distributions. Hence, models of intermediate complexity resulted in the best trade-off to predict species distributions across time. Reliable temporal model transferability is especially relevant for forecasting species distributions under future climate change. Consequently, species-specific model tuning should be used to find the best modeling settings to control for complexity, notably with paleoecological data to independently validate model projections. For cross-temporal projections of species distributions for which paleoecological data is not available, models of intermediate complexity should be selected.

Characterisation of Bolivian savanna ecosystems by their modern pollen rain and implications for fossil pollen records

The majority of vegetation reconstructions from the Neotropics are derived from fossil pollen records extracted from lake sediments. However, the interpretation of these records is restricted by limited knowledge of the contemporary relationships between the vegetation and pollen rain of Neotropical ecosystems, especially for more open vegetation such as savannas. This research aims to improve the interpretation of these records by investigating the vegetation and modern pollen rain of different savanna ecosystems in Bolivia using vegetation inventories, artificial pollen traps and surface lake sediments. Two types of savanna were studied, upland savannas (cerrado), occurring on well drained soils, and seasonally-inundated savannas occurring on seasonally water-logged soils. Quantitative vegetation data are used to identify taxa that are floristically important in the different savanna types and to allow modern pollen/vegetation ratios to be calculated. Artificial pollen traps from the upland savanna site are dominated by Moraceae (35%), Poaceae (30%), Alchornea (6%) and Cecropia (4%). The two seasonally-inundated savanna sites are dominated by Moraceae (37%), Poaceae (20%), Alchornea (8%) and Cecropia (7%), and Moraceae (25%), Cyperaceae (22%), Poaceae (19%) and Cecropia (9%), respectively. The modern pollen rain of seasonally-inundated savannas from surface lake sediments is dominated by Cyperaceae (35%), Poaceae (33%), Moraceae (9%) and Asteraceae (5%). Upland and seasonally-flooded savannas were found to be only subtly distinct from each other palynologically. All sites have a high proportion of Moraceae pollen due to effective wind dispersal of this pollen type from areas of evergreen forest close to the study sites. Modern pollen/vegetation ratios show that many key woody plant taxa are absent/under-represented in the modern pollen rain (e.g., Caryocar and Tabebuia). The lower-than-expected percentages of Poaceae pollen, and the scarcity of savanna indicators, in the modern pollen rain of these ecosystems mean that savannas could potentially be overlooked in fossil pollen records without consideration of the full pollen spectrum available.

Differentiation between Neotropical rainforest, dry forest and savannah ecosystems by their modern pollen spectra and implications for the fossil pollen record

Accurate differentiation between tropical forest and savannah ecosystems in the fossil pollen record is hampered by the combination of: i) poor taxonomic resolution in pollen identification, and ii) the high species diversity of many lowland tropical families, i.e. with many different growth forms living in numerous environmental settings. These barriers to interpreting the fossil record hinder our understanding of the past distributions of different Neotropical ecosystems and consequently cloud our knowledge of past climatic, biodiversity and carbon storage patterns. Modern pollen studies facilitate an improved understanding of how ecosystems are represented by the pollen their plants produce and therefore aid interpretation of fossil pollen records. To understand how to differentiate ecosystems palynologically, it is essential that a consistent sampling method is used across ecosystems. However, to date, modern pollen studies from tropical South America have employed a variety of methodologies (e.g. pollen traps, moss polsters, soil samples). In this paper, we present the first modern pollen study from the Neotropics to examine the modern pollen rain from moist evergreen tropical forest (METF), semi-deciduous dry tropical forest (SDTF) and wooded savannah (cerradão) using a consistent sampling methodology (pollen traps). Pollen rain was sampled annually in September for the years 1999–2001 from within permanent vegetation study plots in, or near, the Noel Kempff Mercado National Park (NKMNP), Bolivia. Comparison of the modern pollen rain within these plots with detailed floristic inventories allowed estimates of the relative pollen productivity and dispersal for individual taxa to be made (% pollen/% vegetation or ‘p/v’). The applicability of these data to interpreting fossil records from lake sediments was then explored by comparison with pollen assemblages obtained from five lake surface samples.

Araucaria forest dynamics in relation to fire frequency in southern Brazil based on fossil and modern pollen dataf

To elucidate the relationship between forest dynamics and fire frequency pollen percentages and charcoal amounts from a 120 cm long peat core and from samples of modern pollen rain were collected along a transect. The study site in southern Brazil is characterized by a species-rich mosaic of grassland-Araucaria forest. It is of crucial importance for management strategies for conservation to understand the development and maintenance of these vegetation mosaics including their sharp forest-grassland boundaries. During the late Holocene, considerable changes occurred in the area. From Anno Domini (AD) 1360 to 1410, the area was dominated by Campos (grassland) vegetation and fire was very common. From AD 1410 to 1500, Araucaria forest expanded and fire was less frequent. From AD 1500 to 1580, Campos grassland spread and the Araucaria forest ceased its development, apparently due to the increase of fire. From AD 1580 to 1935, after a decrease in fire frequency, Araucaria forest expanded again. From AD 1935 to the present, the Araucaria forest expanded while the Campos area decreased. Fire was very rare in this period. The results indicate a strong interaction of forest expansion, forming a mosaic of Campos and Araucaria forest, and the frequency of fire during the past 600 years. A possible collapse of the indigenous population following the post-Colombian colonization in southern Brazil after about AD 1550 may have caused a great reduction of fire frequency. The introduction of cattle (probably after AD 1780) and the resulting decrease of fire frequency might be the reason for forest expansion. Fire is probably the most important factor controlling the dynamics of the forest-grassland mosaics and the formation of sharp borders between these two vegetation types. (C) 2010 Elsevier B.V. All rights reserved.

Pollen percentage thresholds of Abies alba based on 13-year annual records of pollen deposition in modified Tauber traps: perspectives of application to fossil situations

Abies alba (fir), a submontane tree from Central European mountains and uplands, is of special interest for palaeoecological and palaeoclimate interpretations due to its sensitivity to air and soil humidity. Its present distribution limit in the uplands of SE Poland is still a matter of debate. In the Holocene fir expanded to Poland very late, but early fir populations are supposed to occur in the Šumava Mts (Czech Republic). The study aims: to estimate pollen thresholds for fir presence/absence in Bohemia (Czech Republic) and Poland on the basis of modified Tauber pollen traps; to use these thresholds for tracing fir presence in two pollen diagrams from Poland (Słone and Bezedna lakes) in the border zone between the Roztocze region (with fir forest stands today) and Polesie (where fir has never played an important role); and to investigate how the percentage presence/absence threshold can be used to trace the occurrence and abundance of fir trees in the Šumava Mts based on the pollen diagrams of Rokytecká slat' and Mrtvý luh. The fir pollen thresholds estimated in terms of PAR (pollen accumulation rates or pollen influx) range from 843 (grains cm− 2 year− 1) (Roztocze) to 61 (Krkonoše) and 49 (Šumava). Percentage thresholds range from 0.3% in Krkonoše where fir trees are not present within 4 km to 22% in fir-dominated woodland of the Roztocze, providing evidence of strong underrepresentation of fir in the pollen deposition. Application of these percentage thresholds to the Słone and Bezedna pollen diagrams indicates that occurrence of fir in the region is possible from 3.5 cal ky BP onwards, though the evidence is not decisive. In the Šumava, a low representation of fir pollen (1–2%) reflecting presence of scattered fir trees was detected as early as ca. 7.0 cal ky BP.

On trend estimation under monotone Gaussian subordination with long-memory: application to fossil pollen series

Fossil pollen data from stratigraphic cores are irregularly spaced in time due to non-linear age-depth relations. Moreover, their marginal distributions may vary over time. We address these features in a nonparametric regression model with errors that are monotone transformations of a latent continuous-time Gaussian process Z(T). Although Z(T) is unobserved, due to monotonicity, under suitable regularity conditions, it can be recovered facilitating further computations such as estimation of the long-memory parameter and the Hermite coefficients. The estimation of Z(T) itself involves estimation of the marginal distribution function of the regression errors. These issues are considered in proposing a plug-in algorithm for optimal bandwidth selection and construction of confidence bands for the trend function. Some high-resolution time series of pollen records from Lago di Origlio in Switzerland, which go back ca. 20,000 years are used to illustrate the methods.

Pinus nigra (European black pine) as the dominant species of the last glacial pinewoods in south-western to central Iberia: a morphological study of modern and fossil pollen

Aim Our aim was to discriminate different species of Pinus via pollen analysis in order to assess the responses of particular pine species to orbital and millennial-scale climate changes, particularly during the last glacial period. Location Modern pollen grains were collected from current pine populations along transects from the Pyrenees to southern Iberia and the Balearic Islands. Fossil pine pollen was recovered from the south-western Iberian margin core MD95-2042. Methods We measured a set of morphological traits of modern pollen from the Iberian pine species Pinus nigra, P. sylvestris, P. halepensis, P. pinea and P. pinaster and of fossil pine pollen from selected samples of the last glacial period and the early to mid-Holocene. Classification and regression tree (CART) analysis was used to establish a model from the modern dataset that discriminates pollen from the different pine species and allows identification of fossil pine pollen at the species level. Results The CART model was effective in separating pollen of P. nigra and P. sylvestris from that of the Mediterranean pine group (P. halepensis, P. pinea and P. pinaster). The pollen of Pinus nigra diverged from that of P. sylvestris by having a more flattened corpus. Predictions using this model suggested that fossil pine pollen is mainly from P. nigra in all the samples analysed. Pinus sylvestris was more abundant in samples from Greenland stadials than Heinrich stadials, whereas Mediterranean pines increased in samples from Greenland interstadials and during the early to mid-Holocene. Main conclusions Morphological parameters can be successfully used to increase the taxonomic resolution of fossil pine pollen at the species level for the highland pines (P. nigra and P. sylvestris) and at the group of species level for the Mediterranean pines. Our study indicates that P. nigra was the dominant component of the last glacial south-western/central Iberian pinewoods, although the species composition of these woodlands varied in response to abrupt climate changes.

Younger Dryas and Allerød summer temperatures at Gerzensee (Switzerland) inferred from fossil pollen and cladoceran assemblages

Linear- and unimodal-based inference models for mean summer temperatures (partial least squares, weighted averaging, and weighted averaging partial least squares models) were applied to a high-resolution pollen and cladoceran stratigraphy from Gerzensee, Switzerland. The time-window of investigation included the Allerød, the Younger Dryas, and the Preboreal. Characteristic major and minor oscillations in the oxygen-isotope stratigraphy, such as the Gerzensee oscillation, the onset and end of the Younger Dryas stadial, and the Preboreal oscillation, were identified by isotope analysis of bulk-sediment carbonates of the same core and were used as independent indicators for hemispheric or global scale climatic change. In general, the pollen-inferred mean summer temperature reconstruction using all three inference models follows the oxygen-isotope curve more closely than the cladoceran curve. The cladoceran-inferred reconstruction suggests generally warmer summers than the pollen-based reconstructions, which may be an effect of terrestrial vegetation not being in equilibrium with climate due to migrational lags during the Late Glacial and early Holocene. Allerød summer temperatures range between 11 and 12°C based on pollen, whereas the cladoceran-inferred temperatures lie between 11 and 13°C. Pollen and cladocera-inferred reconstructions both suggest a drop to 9–10°C at the beginning of the Younger Dryas. Although the Allerød–Younger Dryas transition lasted 150–160 years in the oxygen-isotope stratigraphy, the pollen-inferred cooling took 180–190 years and the cladoceran-inferred cooling lasted 250–260 years. The pollen-inferred summer temperature rise to 11.5–12°C at the transition from the Younger Dryas to the Preboreal preceded the oxygen-isotope signal by several decades, whereas the cladoceran-inferred warming lagged. Major discrepancies between the pollen- and cladoceran-inference models are observed for the Preboreal, where the cladoceran-inference model suggests mean summer temperatures of up to 14–15°C. Both pollen- and cladoceran-inferred reconstructions suggest a cooling that may be related to the Gerzensee oscillation, but there is no evidence for a cooling synchronous with the Preboreal oscillation as recorded in the oxygen-isotope record. For the Gerzensee oscillation the inferred cooling was ca. 1 and 0.5°C based on pollen and cladocera, respectively, which lies well within the inherent prediction errors of the inference models.

The palaeolimnology of Soppensee (Central Switzerlabnd), as evidenced by diatom, pollen, and fossil-pigment analyses

The development of Soppensee (Central Switzerland, 596 m a.s.l.) has been reconstructed using algal remains such as diatoms, chlorophytes and fossil pigments, as well as the pollen and spores of macrophytes. Sediment accumulation in Soppensee began at the end of the last glacial period, approximately 15,000 yrs ago. During the Oldest Dryas biozone (> 12,700 radiocarbon yrs B.P.) the lake had low primary productivity. After reforestation with birch and later pine, around 12,700 B.P., phases of summer anoxia occurred in the lake. These anoxic conditions were most probably caused by additional carbon input from the catchment, as well as longer phases of stratification due to reduced wind exposure caused by the sheltering effect of increased tree cover. From the Younger Dryas biozone (10,800 to 10,000 radiocarbon yrs B.P.) onwards, Soppensee became meromictic for several millennia.

Fossil pollen record of composite sediment core TMD from Tso Moriri, analysis of modern surface pollen samples, mean annual precipitation reconstruction, and digitisation and recalibration of different discussed palaeoclimate proxy records

This paper presents a new fossil pollen record from Tso Moriri (32°54'N, 78°19'E, 4512 m a.s.l.) and seeks to reconstruct changes in mean annual precipitation (MAP) during the last 12,000 years. This high-alpine lake occupies an area of 140 km**2 in a glacial-tectonic valley in the northwestern Himalaya. The region has a cold climate, with a MAP <300 mm, and open vegetation. The hydrology is controlled by the Indian Summer Monsoon (ISM), but winter westerly-associated precipitation also affects the regional water balance. Results indicate that precipitation levels varied significantly during the Holocene. After a rapid increase in MAP, a phase of maximum humidity was reached between ca. 11 to 9.6 cal ka BP, followed by a gradual decline in MAP. This trend parallels the reduction in the Northern Hemisphere summer insolation. Comparison of different palaeoclimate proxy records reveal evidence for a stronger Holocene decrease in precipitation in the northern versus the southern parts of the ISM domain. The long-term trend of ISM weakening is overlaid with several short periods of greater dryness, which are broadly synchronous with the North Atlantic cold spells, suggesting reduced amounts of westerly-associated winter precipitation. Compared to the mid and late Holocene, it appears that westerlies had a greater influence on the western parts of the ISM domain during the early Holocene. During this period, the westerly-associated summer precipitation belt was positioned at Mediterranean latitudes and amplified the ISM-derived precipitation. The Tso Moriri pollen record and moisture reconstructions also suggest that changes in climatic conditions affected the ancient Harappan Civilisation, which flourished in the greater Indus Valley from approximately 5.2 to 3 cal ka BP. The prolonged Holocene trend towards aridity, punctuated by an interval of increased dryness (between ca. 4.5 to 4.3 cal ka BP), may have pushed the Mature Harappan urban settlements (between ca. 4.5 to 3.9 cal ka BP) to develop more efficient agricultural practices to deal with the increasingly acute water shortages. The amplified aridity associated with North Atlantic cooling between ca. 4 to 3.6 and around 3.2 cal ka BP further hindered local agriculture, possibly causing the deurbanisation that occurred from ca. 3.9 cal ka BP and eventual collapse of the Harappan Civilisation between ca. 3.5 to 3 cal ka BP.