Colombian biomes are reconstructed at 45 sites from the modern period extending to the Last Glacial Maximum (LGM). The basis for our reconstruction is pollen data assigned to plant functional types and biomes at six 3000-yr intervals. A reconstruction of modern biomes is used to check the treatment of the modern pollen data set against a map of potential vegetation. This allows the biomes reconstructed at past periods to be assessed relative to the modern situation. This process also provides a check on the a priori assignment of pollen taxa to plant functional types and biomes. For the majority of the sites, the pollen data accurately reflect the potential vegetation, even though much of the original vegetation has been transformed by agricultural practices. At 18 000 ¹⁴C yr BP, a generally cool and dry environment is reflected in biome, assignments of cold mixed forests, cool evergreen forests and cool grassland–shrub; the latter extending to lower altitudes than presently recorded. This signal is strongly recorded at 15 000 and 12 000 ¹⁴C yr BP, the vegetation at these times also reflecting a relatively cool and dry environment. At 9000 ¹⁴C yr BP there is a shift to biomes thought to result from slightly cooler environmental conditions. This trend is reversed by 6000 ¹⁴C yr BP; most sites, within a range of different environmental settings, recording a shift to more xeric biome types. There is an expansion of steppe and cool mixed-forest biomes, replacing tropical dry forest and cool grassland–shrub biomes, respectively. These changes in biome assignments from the modern situation can be interpreted as a biotic response to mid-Holocene climatic aridity. At 3000 ¹⁴C yr BP the shift is mainly to biomes characteristic of slightly more mesic environmental conditions.
Radiocarbon dates from two sites in the Andes (Ecuador and Peru) confirm that glaciers culminated a readvance after 11 000 yr BP. A moraine stage, equivalent in altitude and position relative to existing glaciers, is present in most glacierized ranges, but its age is equivocal. Broadly limiting dates from Colombia and Peru suggest that the stage may be Late-glacial, as it is younger than 12100 yr BP, but formed before the early Holocene; in southern Chile a comparable moraine stage is older than ca. 9100 yr BP. Andean glaciers appear to have advanced at least twice during the Late-glacial interval.Glacier reconstruction from these moraine limits suggests depression of the equilibrium line altitude by at least 300–400 m in the northern and north-central Andes, and possibly less than this farther south.Late-glacial climatic change occurred globally and possibly reflects North Atlantic temperature and circulation changes forced by deglaciation of the northern ice sheets, migrations north and south of the Atlantic Polar Front, and the switching off and on of a ‘dust pump’ in low midlatitudes.
Monthly temperature and precipitation estimates for the period 15000–2000 yr BP are statistically derived from pollen analytical data obtained from a site near Echternach, Luxembourg. A continuous warming trend of over 10°C is recorded for the period 10500—8000 yr BP, along with a rise in precipitation of 500 mm. A fairly constant temperature characterised the period between 8000 and 2000 yr BP, though an episode of slight temperature decrease (-2°C) occurred at about 4500 yr BP. The most humid intervals seem to have been at around 9000–7000 and 5000–3000 yr BP (principally during the warm season: more continental influence). In contrast with results obtained from more southerly sites, the ‘Older Dryas’ climatic reversal is clearly recorded at Echternach slightly before 12000 yr BP. It is marked by a cooling in July temperature of about 4°C and a decrease in monthly precipitation of about 30 mm (40%).
A hypothesis is presented that late Quaternary 100 000-yr glacial cycles are driven by an asymmetric thermohaline–ice-sheet oscillator that emerged in the global climate system 650 000–950 000 yr ago, perhaps when the main source of Northern Hemisphere deep-water production shifted south from the Arctic into the Nordic seas. It is hypothesised that the asymmetry is due to the increasing difficulty after 950 000 years ago of resetting an interglacial mode of the critical Nordic limb of the salinity conveyor once it switches off and an ensuing iceberg flux enters the areas of downwelling. A possible reason for both a southward shift and the resulting asymmetry is uplift of the Greenland–Scotland submarine ridge from activity of the Iceland mantle plume.
Pollen data are the most important source of information with respect to late Quaternary vegetation history. Broad-scale palaeovegetation patterns have been subjectively inferred from mapped pollen data by previous authors. In this study, multivariate classification was applied to European pollen data for the last 13 000 yr. The resulting clusters are mapped at millennial intervals and can be equated with vegetation units. The maps portray the changing vegetation of Europe since the last glacial. They reveal the impermanence of the assemblages of species that ecologists recognise as communities. The dominant patterns in the maps also change through time, indicating important changes in palaeoenvironmental conditions and in the alignment of major environmental gradients. Human impact upon European vegetation history is seen to be relatively unimportant when the vegetation is viewed at a continental scale.
The Late Quaternary glaciation of Tibet has received considerable attention in the last few decades due to its influence on the regional climate, especially the Asian summer monsoon. Recently, however, it has been argued that the Tibetan ice sheet also might have played an important role in initiating global-scale palaeoclimatic changes. Controversy, however, exists on the nature of Late Quaternary ice cover over Tibet due largely to the subjectivity in the interpretation of the sparse and complex geomorphological evidence. We have examined this problem in the light of δ 18O data (a temperature proxy) of ice cores-from the Dunde ice cap on the northern flank of Tibet. Considering only the gross features in the Dunde ice-core isotopic data, we have interpreted a temperature decrease of 4°–6°C and consequent lowering of equilibrium line altitude (ELA) in the range 700–850 m during the last glacial stage (LGS). This could have caused depression of the snow line below the mean altitutde of the Tibetan plateau, resulting in an areally extensive but marginally thick ice cover. However, if one also considers the possibility that precipitation on the Tibetan plateau during LGS may have been significantly lower than at present, the ELA depression would be much less than that estimated by considering the temperature effect alone.
Pollen and macrofossil analyses of a sediment core from Beaver Pond (60° 37' 14″ N, 154° 19' W, 579m a.s.l.) reveal a record of regional and local postglacial vegetation change in south-western Alaska. The chronology is based on five AMS (accelerator mass spectrometry) 14C ages obtained from terrestrial plant macrofossils. Pollen and macrofossil records suggest that open herb and shrub tundra with e.g. Poaceae, Cyperaceae, Artemisia, Vaccinium and Salix prevailed on the landscape before ca. 14000 cal a BP. The shift from herb- to shrub-dominated tundra (Salix, subsequent Betula expansion) possibly reflects climatic warming at the beginning of the Bølling period at ca. 14700-14500 and around 13500 cal a BP. Vegetation (Betula shrub tundra) remained relatively stable until the early Holocene. Macrofossil influx estimates provide evidence for greater biomass in Betula shrub tundra during the early postglacial period than today. Charcoal accumulation rates suggest tundra fire activity was probably greater from ca. 12500 to 10500 cal a BP, similar to results from elsewhere in Alaska. The pollen and macrofossil records of Beaver Pond suggest the prevalence of low shrub tundra (shrub Betula, Betula nana, Vaccinium, Ledum palustre, Ericaceae) and tall shrub tundra (Alnus viridis ssp. crispa, Salix) between 10000 and 4000 cal a BP. This Holocene vegetation type is comparable with that of the modern treeless wet and moist tundra in south-western Alaska. The expansion of Picea glauca occurred ∼4000 cal a BP, much later than that of A. viridis (ssp. crispa), whereas in central and eastern Alaska Picea glauca expanded prior to or coincident with Alnus (viridis). At sites located only 200-400km north-east of Beaver Pond (Farewell and Wien lakes), Picea glauca and Betula forests expanded 8000-6000 cal a BP. Unfavourable climatic conditions and soil properties may have inhibited the expansion and establishment of Picea across south-west Alaska during the mid and late Holocene.
A synthesis of the main environmental changes that are interpreted to have occurred in Iceland during the period 13-9 ka BP is presented. Most of the evidence available relates to variations in the position of ice margins, although some limited information on vegetation history and soil stability is also referred to. Only qualitative and limited climatic inferences can be made for this period because of the lack of detailed evidence. A summary curve of the relative extent of ice cover is presented.
A pollen diagram from Lago di Martignano, a maar lake in central Italy, provides an 11000-year record of vegetation and environment change. The earliest pollen spectra are dominated by Artemisia and Gramineae, representing late glacial steppe vegetation typical of the Mediterranean region. Broad-leaved forests were established by ca. 11 000 yr BP. Although Quercus initially dominated their canopy, a wide range of other mesophyllous trees were also present. Pollen values for sclerophyllous tree and shrub taxa characteristic of Mediterranean woodlands and scrub are initially low (<10%). After ca. 7000 yr BP, however, they begin to increase and rise to a peak of >40% of total land pollen at ca. 6700 yr BP, with Olea europaea the single most abundant taxon. Human influence upon the vegetation only becomes significant somewhat after this peak, with progressive clearance of woodland and expansion of herbaceous communities. Castanea sativa and luglans regia pollen is recorded consistently from the beginning of the rise in pollen values for taxa characteristic of Mediterranean scrub communities. Pollen values for arable crops increase progressively after ca. 5500 yr BP, following the peak pollen values for taxa characteristic of Mediterranean scrub vegetation. Late glacial and Holocene climate changes have been complex in this region, with the present character of the climate developing only during the last millennium. Rates of change of pollen spectra peak during this period.
The core interval of 0–140 m of the 357-m-long pollen record from site Funza-I was analysed using multivariate analysis. This long continental record is from the high plain of Bogotá, located at 2550 m altitude in the Eastern Cordillera of Colombia. Time control is based on a visual match with core Funza-II from the same location, which includes absolute fission-track ages, and a match with the δ¹⁸O record of ODP Site 677. We reassessed the time-frame of Funza-I compared with previous papers, and conclude that the period of ca. 650–25 kyr BP is represented.