The Holocene

Reconstructing past fire requires ways to compare concentration estimates obtained by different charcoal analysis methods. These methods estimate concentrations from area measurements or particle counts on pollen slides, on sieves, and on thin sections, and they do so at different magnifications. Depending on particle properties, each method might give substantially different results. Image analysis of charred particles in thin sections of lake sediments is used to parameterize relationships among different concentration estimators. We determined distributions of particle size, shape, and orientation in undisturbed sediments and assessed how those distributions affect concentration estimates. Particle-diameter distributions were conservative across different magnifications. Particle shape was also conservative, with major:minor axis ratios near 2.5. Our derived relationships among estimators could be used to improve comparability for all but the smallest particle classes, which tended to be unpredictable in orientation. Simple equations with nomograms for necessary parameter values are provided. We derive a settling velocity index for composite samples that can be used to estimate relative differences in source region and dispersion. Comparisons of estimates from lakes in eastern North America showed geographic coherency, with highest values in the midwest and low values in the nor theast. A single exception to the eastwardly decline came from a sand plain site in New Hampshire, where fires were expected to be more common than in the prevailing northern hardwoods of the region. Application of our method to previous estimates of mass fluxes in eastern North America show values closer to those expected based on current understanding of fire behaviour than were previous, uncorrected estimates.

Increased aridity throughout the Great Plains and Rocky Mountain region during the middle Holo cene has been documented from pollen records, aeolian proxy variables in lake cores, and active sand dune migration. Varve calibration provided by a continuously varved record of the Holocene from a core from Elk Lake, northwestern Minnesota, shows that the influx of aeolian clastic material increased beginning about 8 ka and ended about 3.8 ka, with peak aeolian activity at about 6 ka. If aeolian influx to Elk Lake corresponds in time to aeolian influx in other lakes and to maximum dune activity in Minnesota dune fields, then the varve calibration in Elk Lake provides precise time calibration of periods of peak aeolian activity in Minnesota. Palaeowind studies from the Minnesota dune fields show that the dominant wind direction when the dunes were active was from the northwest, the same as the dominant wind direction in dune fields throughout the Great Plains and Rocky Mountains. If the mid-Holocene aeolian activity in Minnesota was driven by an increase in westerly zonal winds, then the varve calibration can be extended to more precisely determine the timing of activity of dunes over a much broader area. We suggest that an increase in the westerly zonal wind field might have a solar-geomagnetic cause.

The 8500- to 7000-year-old shell-bearing sediments in shallow claypans near Wadi Mansurab in north-central Sudan appear to reflect a widespread and much wetter period in northeast Africa. Stable carbon and oxygen isotope analysis of the Wadi Mansurab gastropods reveals that the shell carbonate is highly depleted in ¹³C and ¹⁸O. The highly depleted oxygen isotopic composition is indicative of rainfall derived from a distant oceanic source. The highly negative nature of the oxygen isotopes indicates that the region had significantly less evaporation than today while the extreme variability in isotopic composition (up to 6-7‰ PDB) is consist ent with a seasonal rainfall regime, characterized by a high degree of inter-annual variability. The dominance of semi-aquatic and swamp-dwelling gastropods over truly aquatic species implies that the region was probably a seasonally flooded grassplain similar to the toich-lands of south-central Sudan today. Our data support the inference that towards 8500-7000 BP there was a stronger southwest monsoon and an associated northward shift in the summer rainfall zone, which caused the Wadi Mansurab region in north-central Sudan to be season ally flooded during a wetter and possibly cooler period with lower rates of evapotranspiration. From 8500 to 7000 BP, lake levels were high elsewhere in northern Africa, suggesting a regionally wetter climate at that time.

Stratigraphical analyses based on 1028 dated Holocene peat cores were used to estimate the rate of carbon accumulation in Finnish peatlands. Results were compared with data from other Boreal areas. The basal age of peat columns was found to be the best predictor of carbon accumulation and a significant correlation between depth and age of peat was evident. When normalized for the mean depth of peatlands in Finland (c. 1.5 m), the average long-term carbon accumulation was 26.1 g C m^-2 yr^-1. In individual cores the values ranged from 2.8 to 88.6 g (average 22.5 ± 11.6 g) being much higher in bogs than in fens and almost double in southern mires as compared with those in the northern Boreal zone. The modelled rate of actual carbon accumulation is about 2/3 of that above. Boreal mires are actual sinks of carbon for thousands of years ahead in the present climate. The predicted greenhouse warming may shift the present Sphagnum mires northward and the net effect may be to restrain the radiative forcing. Poorly known 3D mass accumulation in mires and unpredictable functioning of northern mire ecosystems to climatic changes (for example via increased mire fires) complicate this interpretation.

Lithofacies and benthic foraminiferal analyses of two sediment cores (BS1191-K13B and K14) from Nansen Fjord, eastern Greenland, show evidence of changing oceanographic and sea-ice conditions from AD 730 to the present. Radiocarbon dates on foraminifera provide century-scale resolution of events, but the frequency of events between dated levels can be estimated with a resolution of approximately 4 to 9 years per centimetre of core. Unstratified glacial-marine diamicton with a calcareous foraminiferal fauna gives way to interstratified diamicton and mud with a predominantly agglutinated foraminifera. The diamictons represent periods of season ally open water with strong continuous iceberg rafting. Mud layers were deposited during intervals of prolonged sea-ice cover. The evidence suggests that the climate in the region of Nansen Fjord was warmer and more stable than today during a 'Medieval Warm Period' between c. AD 730 to 1100. Variable climatic conditions with frequent intervals of severe cold characterize a 'Little Ice Age' type interval from c. AD 1630 to 1900. An earlier cold interval culminated c. AD 1370. The record is similar to the 1000-yr-long Icelandic sea-ice record and, to a lesser extent, to the central Greenland Crete ice-core record.

Pollen records from two lake sites located in areas of modern heathland were used to reconstruct the local development of this vegetation type. A pollen record obtained from Loch of Knitchen on Rousay covered the last 6000 radiocarbon years. Betula-Corylus scrub was once locally present, although never a dominant community. Heath vegetation began to develop around 5700 BP. At about 5400 BP woodland taxa declined, possibly in response to Neolithic human activity, and heathland expanded. Calluna vulgaris increas ingly dominated the record thereafter, reflecting local blanket peat development. Data from Loch of Torness on Hoy cover the period from 7700 BP to around 4800 BP. Initial tall herb-type vegetation was replaced by a mixture of trees and heath taxa at around 7000 BP. Arboreal taxa declined at about 6000 BP, and the sub sequent record showed the development of maritime-type heath. The data show that heathland vegetation has been present in Orkney since at least 7000 BP, and that a range of causal processes have been important.

Lake Gosciąz is located in the zone of the maximum extent of the last ice sheet. Distinct lake terraces and deposits related to the lateglacial transgression do not exist. Detailed examination of annually laminated sediments and landforms revealed three phases of higher water level and storage, represented by shore features, by littoral, deltaic and bog deposits, and by higher rates of lacustrine deposition in the pelagic zone. Holocene lake transgressions at c. 8.4-7.7 ka, 2.5-2.3 ka, and less distinct at 5-4 ka BP coincide with phases of higher flood frequency in Poland and glacial advances in the Alps. Because of sedimentation, isolation from groundwaters, encroachment by peatland and drainage of the lake, lake volume during the last 12 ka diminished by as much as 70%, and current water-level fluctuations of about 1-2 m can change water storage by 20-25%.

Regional pollen diagrams indicate that Fagus may have been present close to its northern distri butional limits in southern Sweden for as long as 1000 years. Fagus has probably always been rare in this region, except for some small and scattered stands. Many of these outlying Fagus stands have been regarded as planted but no previous studies have investigated this in detail. At Mattarp, a small outpost Fagus stand in northern Småland, the pollen record of the deposit of a small hollow indicates that Fagus became established c. 400 BP. The local establishment was associated with an earlier phase of woodland clearance and the former system of land-use. Local factors seem to have controlled the immigration and establishment at the stand-scale at this site. Regional climate had probably been favourable for Fagus for almost 2000 years and cannot be regarded as the direct cause for establishment at this site. The complex south Swedish cultural landscape may have favoured Fagus establishment in many areas, particularly when there has been a temporary reduction in human influence. Picea became regionally established around 800 BP, but a local expansion did not occur until c. 400 BP, coinciding with the establishment of Fagus. Picea expansion also seems to have been favoured by a moderate human influence at this site. Fagus pollen influx values have steadily increased since the time of establishment. It seems likely that the Fagus stand has continually increased in size, and probably would continue to do so if this were not prevented by human activities.

Magnetic minerals found in sediments can either be transported to the basin from the catchment area or can be authigenically formed in the sediments. SEM studies, x-ray diffraction spectra and magnetic mineralogical analyses performed on magnetic extracts from freeze-dried sediments of Lake St Moritz, Switzer land, undoubtedly identified the presence of greigite. The occurrence of this magnetic iron sulphide (Fe₃S₄) is associated with changing environmental conditions that can be interpreted from historical data available for the area. Results of the combined XRD and SEM study show that authigenically formed greigite in Lake St Moritz sediments displays a predominant framboidal crystal arrangement. Amorphous, irregular shapes and spherical aggregates, most often associated with this mineral, are present in lower concentration. These results provide a natural example of framboidal iron sulphide formation and support previous experimental results that have shown that oxidation state and iron availability are critical factors for greigite formation. Despite its metastable character, greigite can also be preserved in the geological record. Its magnetic signature may, in some cases, be representative of a well-defined depositional environment.

This paper describes a technique for the enumeration of tephra shards from lake sediment cores whereby approximately 85% of unwanted sediment fractions (organic matter, biogenic silica, carbonates) are removed by selective stepwise chemical attack. The order of 100 samples can be prepared simultaneously in a modest-sized water-bath and the recovery rate of the technique is greater than 90%. The procedure is appli cable to all lake sediment types.

This study reports new data concerning the Holocene glaciation of Ameghino Glacier, an eastern outlet of the Southern Patagonia Icefield. The lowest 7 km of the valley occupied by Ameghino Glacier is now ice-free, of which the upper 3.7 km is a moraine-dammed proglacial lake. Many glacial features are present in the Ameghino Valley. Across the valley floor, four terminal moraines have been identified, each representing a Neoglacial advance. Prominent vegetation trimlines are associated with the youngest of these moraines (M IV) which dammed the proglacial lake. Samples for ¹⁴C dating collected at the trimlines and from the valley floors indicate that the formation of M IV, the culmination of the recent maximum glaciation, occurred c. AD 1600-1700, and that deposition of the valley train occurred c. AD 1500. Large standing trees which were buried and killed by this period of deposition have subsequently been exhumed by river erosion. The glacier may have remained at the trimline level until around AD 1870-1880, the date obtained by Nichols and Miller (1951) from dendrochronological analyses. The number of Neoglacial advances (4) identified in the Ameghino Valley is the same as that recently identified at two other Patagonian glaciers, Upsala and Tyndall, and the date of the latest glacial episode, the 'Little Ice Age', is found to be synchronous. However, the ages of the earlier three Neoglacial episodes remain unknown.