Info

Puerto del Hambre

FIGURE 3 (continued) CI.

lacks a modern vegetation analog. It is interpreted as indicating high precipitation and frost-free climatic conditions. The end of this interval is uncertain.

In the Rocky Mountains, at the site of Mary Jane in Colorado (9), deglaciation began after 17,000 cal. B.P. A mixture of cold-adapted tundra beetles with northern boreal/ subalpine vegetation elements is recorded, suggesting a lowering of the tree line by ca. 500 m and representing a cooler climate. In records from Yellowstone and Grand Teton National Parks (11), an increase of Picea-Pinus assemblages is recorded between 15,300 and 14,800 cal. B.P., attesting to wetter climatic conditions. In Yosemite National Park, the Swamp Lake (10) record shows vegetation composed of Pinus, Tsuga, and Abies between 16,400 and 12,300 cal. B.P. This assemblage has no modern analog and is interpreted as an enriched mixed forest, attesting to cooler temperatures and an increase in precipitation. Wetter climatic conditions are also interpreted from paleolake levels (Thompson et al., 1993; Webb et al., 1993) and packrat middens, indicating greater seasonality (Betancourt et al., 1990).

At Chalco Lake in Mexico (14), between 17,500 and

FIGURE 4 Main vegetation types and climatological information for the interval 15,500-14,500 cal. B.P.

precedent interval no record G hiatus + beginning of the record ^ warmer ^ cooler O drier O wetter >1000 m

FIGURE 5 Main vegetation types and climatological information for the interval 14,500-12,700 cal. B.P.

precedent interval no record G hiatus + beginning of the record ^ warmer ^ cooler O drier O wetter >1000 m

FIGURE 6 Main vegetation types and climatological information for the interval 12,700-11,000 cal. B.P.

FIGURE 5 Main vegetation types and climatological information for the interval 14,500-12,700 cal. B.P.

11,500 cal. B.P., Pinus dominated together with meso-phytic forest, including mainly Abies, Liquidambar, Carpinus, and Corylus. This assemblage reflects high-moisture conditions. The 17,500 cal. B.P. date is, in fact, the upper limit of an interval between 17,500 and 14,500 cal. B.P. At La Yeguada in Panama (15), between 17,000 and 14,700 cal. B.P., moist montane forest is recorded, suggesting humid climatic conditions.

In the high Andean cordillera, four sites were examined: two in Colombia, one in Peru, and one in Venezuela. In the Pedro Palo, Colombia, record (16), between 15,500 and 14,000 cal. B.P., the dominance of nonarboreal pollen suggests dry climatic conditions. Laguna Ciega, Colombia (17), records an increase in Asteraceae between 15,000 and 14,500 cal. B.P. that indicates cold climatic conditions. This interval is called the La Ciega interval and has also been recognized in other records in this area. The record for Laguna Baja, Peru (18), starts at ca. 16,000 cal. B.P.; moist and wet montane elements are well represented for that time.

FIGURE 6 Main vegetation types and climatological information for the interval 12,700-11,000 cal. B.P.

Climatic conditions are defined as having been wet, a condition that lasted until ca. 13,500 cal. B.P. In Mu-cubaji, Venezuela (19), deglaciation started ca. 14,800 cal. B.P. Pollen content at the onset attests to the presence of a desert paramo, suggesting cold conditions between 14,800 and 14,200 cal. B.P.

For the tropical lowlands, we discuss four sites. Lake Valencia, Venezuela (20), in the tropical lowlands, records dominance of Alternanthera, Cyperaceae, and Poaceae between 16,700 and 10,500 cal. B.P., suggesting generally arid conditions. For that time, only a minor increase in moisture is indicated by an increase of savanna tree taxa such as Bursera and Spondias. During the late glacial, Lake Pata, Brazil (21), recorded an assemblage of rain forest taxa with Podocarpus for which no modern analogs exist in present-day vegetation. This record was interpreted to indicate cold temperatures and humid climatic conditions. At Carajas, Brazil (22), a gap in sedimentation is recorded until ca. 14,500 cal. B.P. At Salitre, Brazil (23), the presence of Araucaria forest elements between 15,300 and 12,300 cal. B.P. attests to cold and humid climatic conditions.

Five sites are discussed for temperate South America. At Puerto Octay, Chile (25), a gap occurs between ca. 15,500 and 14,500 cal. B.P. At Pichihue, Chile (24), the presence of Nothofagus, a pioneer tree taxon, and moorland taxa between 15,000 and 14,500 cal. B.P. attests to humid climatic conditions. At Puerto del Hambre (26) and Punta Arenas (27), Chile, the presence of an Em-petrum steppe-tundra (heath) is recorded between 16,000 and 14,000 cal. B.P., attesting to drier climatic conditions. At Harberton (28), Argentina, Empetrum heathland is also recorded between 16,000 and 15,000 cal. B.P., indicating an increase in temperature and a decrease in precipitation.

20.4.2. Time Interval Between 14,500 and 12,700 cal. B.P.

A summary of the results for the period 14,50012,700 cal. B.P. is reported in Fig. 5 (the asterisk (*) indicates that no change in vegetation or climate occurred compared to the previous interval).

In Alaska, the Ongivinuk record (1) shows an increase of Betula shrubs in the tundra landscape between ca. 13,000 and 12,700 cal. B.P. This phase is related to a warmer and moister climate compared to the previous interval. The Pleasant Island (2) record shows the presence of a Pinus parkland with Alnus and Salix, characteristic of a wetter and warmer climate.

At the Olympic Peninsula (3), the presence of seeds of Opuntia (Cactaceae) suggests dry and warm summers between 15,000 and 13,000 cal. B.P. At Gordon Lake (5), a closed forest with Abies amabilis and Tsuga heterophylla is recorded between 14,500 and 12,800 cal. B.P. At Little Lake (4), an increase in the frequency of Pseudotsuga is recorded between 14,250 and 12,400 cal. B.P.

Allamuchy and Linsley Pond records (7) show no vegetation change since the previous interval. The same mixed thermophilous deciduous boreal forest with Picea, Larix, Abies, and Quercus is recorded, which suggests a cool and humid climate. In the Jackson Pond record (8), there is also no difference compared to the previous interval, although sampling resolution of one sample for 700 years is too low to detect any oscillation. The climate is defined as having been cool and moist. In the Browns Pond record (12), between ca. 13,000 and 10,000 cal. B.P., a Quercus-Ostrya-Carpinus association is recorded, attesting to warmer and wetter climatic conditions. At Camel Lake (13), Picea disappeared and deciduous tree taxa, Fagus and Quercus, began to dominate between 14,800 and 11,500 cal. B.P. The climate was warmer and wetter than during the previous interval.

In the Rocky Mountains, the Yellowstone and Grand Teton National Park sites (11) record an increase of Picea between 13,500 and 11,500 cal. B.P., attesting to wetter and warmer climatic conditions. In Mary Jane (9), sediment deposition stopped.

On the Pacific coast, pollen assemblages show no change compared to the previous interval and, although no modern analogs exist for this type of rich mixed forest, the climate is defined as cool and wet.

In Central America, no change in vegetation composition occurred at Chalco Lake (14) and La Yeguada (15). In the high Andean cordillera, the Pedro Palo record (16) shows an expansion of open scrub and grass paramo characteristic of a climatic warming between 14,000 and 13,000 cal. B.P. An increase of arboreal pollen is recorded at Laguna Ciega (17) between 14,500 and 13,000 (12,700) cal. B.P., attesting to wetter climatic conditions, particularly between 14,200 and 14,000 cal. B.P. The Laguna Baja record (18) shows no difference in vegetation composition compared to the previous interval.

In the tropical lowland records, at Lake Pata (21), no sediment deposition occurred during this time interval. At Carajas (22), when sedimentation restarts at ca. 14,600 cal. B.P., an increase of arboreal pollen is recorded, attesting to a moist climate until ca. 10,000 cal. B.P. The Salitre record (23) shows no change compared to the previous interval; the climate was still cool and wet.

In southern South America, at Puerto Octay (25), the expansion of North Patagonian rain forest species is indicative of a warming. At Pichihue (24), the development of peat and a succession of different Magellanic tundra taxa characterize a gradual evolution of the edaphic vegetation, which may indicate equally cool but wetter conditions compared to the previous interval.

In the Punta Arenas (27) and Puerto del Hambre (26) records, an increase in Nothofagus is recorded between 14,500 and 12,700 cal. B.P., suggesting a wet interval. In Harberton (28), the increase of Poaceae together with mesic taxa attests to an increase in moisture under continuing cooler temperatures.

20.4.3. Time Interval Between 12,700 and

11,000 cal. B.P., Younger Dryas Chronozone

The data for the time interval 12,700-11,000 cal. B.P. are compiled in Fig. 6. In Alaska, at Ongivinuk and Grandfather Lakes (1), the earlier Betula shrub tundra was replaced by herb tundra with Poaceae and Artemisia, suggesting a climate that was drier and cooler than before.

In Alaska, the Pleasant Island record (2) shows Pinus parkland taxa decreased and instead herb tundra re turned between 12,300 and 11,400 cal. B.P. This change is interpreted as indicating a drier and cooler climate. In the Olympic Peninsula record (3), the increase in Pinus, Picea, and Tsuga recorded between 13,000 and 11,000 cal. B.P. suggests an increase of precipitation. At Gordon Lake (5), the presence of T. mertensiana with Pinus and Abies characterizes a cooler and drier climate between 12,800 and 11,000 cal. B.P. In the Little Lake record (4), between 12,400 and 11,000 cal. B.P., the increase of a mixed coniferous forest with Pinus reversed the expansion of the Pseudotsuga forest, attesting to a cooler climate, although there is a lack of clear modern analogs.

On the northeastern Atlantic coast, at the Killarney site (6), an increase of boreal forest taxa is recorded. The northward migration of deciduous tree species was stopped, and a two-step climate reversal was recorded. The first one, called the Killarney oscillation, occurred between 13,000 and 12,800 cal. B.P.; the second one occurred between 12,700 and 11,000 cal. B.P. In the Alla-muchy and Linsley Pond records (7), there is no change in vegetation composition compared to the previous interval. The climate was still moist and warm. Records from the Yellowstone area do not show any climatic reversal. The climate was wet and cool as in the previous interval. At Browns Pond (12), no reversal is recorded, although the vegetation assemblage of Quercus-Ostrya-Carpinus indicates a drier climate than in the previous interval (new data do show a reversal, D. Peteet, personal communication). At Camel Lake (13), a gap in sedimentation occurred between 12,000 and 11,000 cal. B.P.

In Central America, the Chalco Lake record (14) does not show any change since the previous interval and moist montane forest elements continue. In Costa Rica, at the La Chonta 2 site (not shown), the forest line dropped from 2800-2400 m elevation; this event is interpreted as indicating a 2°-2.5°C cooling (Hooghiem-stra et al., 1992). In the La Yeguada record (15), an increase of tropical lowland taxa attests to disturbances of the rain forest and was interpreted to reflect a wetter climate between 13,000 (13,300-12,700) and 12,300 (12,700-12,000) cal. B.P.

In the high Andean cordillera, the Pedro Palo record (16) shows a two-step reversal. The first step, between 13,000 and 12,300 cal. B.P., is characterized by a decrease of Andean and sub-Andean elements, which is interpreted as indicating a cooling. The second step, between 12,300 and 12,000 cal. B.P., shows an increase in paramo elements, which is interpreted as reflecting drier climatic conditions. At Laguna Ciega (17), an increase in Poaceae, together with a decrease of arboreal pollen taxa, is defined as representing a cold and dry interval called El Abra. In the Mucubaji record (19), superparamo plant assemblages increased between 13,500 and 10,500 cal. B.P., attesting to cooler climatic conditions. At Laguna Baja (18), between 13,500 and 11,000 cal. B.P., moist montane forest elements were replaced by Poaceae, and the charcoal concentration also increased. Three environmental and climatic scenarios could explain this observation: (1) the tree line shifted to lower elevations in response to a cooling; (2) paramo vegetation expanded, implying increased aridity, an explanation that is supported by evidence for increased fires; and (3) arboreal pollen are allochthonous and represent locally an increase in long-distance transport of montane forest pollen.

In the tropical lowlands, there is no sediment deposition during this time interval at Lake Pata (21). At Carajas (22), the vegetation composition did not change since the previous interval. At Salitre (23), a brief increase of Apiaceae and Poaceae between 12,300 and 10,000 cal. B.P. a decrease of the Araucaria forest elements, is interpreted as indicating a shift to drier climatic conditions.

In the Puerto Octay (25) record, a cooling is indicated by a decrease of rain forest taxa and an increase of Nothofagus dombeyi-type, Maytenus disticha, and Podo-carpus nubigena spp. However, at the same time, charcoal particles increased, suggesting that forest fires and not climate could have produced this change in tree taxa. In the Pichihue (24) record, there is no vegetation change. In the Punta Arenas (27), Puerto del Hambre (26), and Harberton (28) records, a Poaceae-Empetrum assemblage suggests drier climatic conditions between 12,900 (12,500) and 11,500 cal. B.P.

20.5. GENERAL DISCUSSION: CHARACTERIZATION OF THE LATE GLACIAL

The establishment of modern forests occurred at different times and followed different patterns, probably related as much to the records' locations as to regional climate change. In spite of differences in resolution and unequal chronological control, the data from the Americas show no clear evidence for synchroneity of climatic oscillations during the late glacial interval. We will point out some of the reasons for this lack of syn-chroneity.

Several vegetation records show no vegetation changes during the intervals considered. In the Pacific Northwest, for instance, the conifer forest with Pinus, Tsuga, and Abies dominated between 16,000 and 14,000 cal. B.P. This forest was replaced by subalpine forest taxa between 14,000 and 11,500 cal. B.P. In eastern North America, a mixed Picea forest existed between

17,000 and 13,500 cal. B.P., after which it was replaced by a deciduous Quercus forest. In southern Central America, the semi-evergreen forest with Abies, Liquid-ambar, Carpinus, Alnus, Corylus, Ulmus, Juglans, Fagus, and Populus was fully developed between 18,500 and 11,000 cal. B.P. In tropical lowlands, rain forest taxa are mixed with mountain forest taxa, such as Podocarpus and Hedyosmum. This mixed forest became fully developed after 14,000 cal. B.P.

However, other vegetation records show several oscillations that are not coeval with circum-North Atlantic oscillations. For example, in the records from the southernmost latitudes, several oscillations between Empetrum heathland and Poaceae steppe-tundra occurred between 16,500 and 13,500 cal. B.P. and 13,500 and 11,000 cal. B.P. Repeated changes in moisture are also recorded for midlatitude records in temperate southern South America, including the onset of peat growth, fluctuations of Magellanic moorland elements, and Nothofagus forest taxa. Although the timing of these late glacial fluctuations is not strictly coeval, interhemi-spheric synchroneity is concluded by some authors (Denton et al., 1999; Moreno et al., 1999; but see the discussion by Markgraf and Bianchi, 1999).

Based on the characteristic patterns of late glacial oscillations, the records discussed previously can be divided into three groups:

1. High-altitude sites and high northern and mid-northern latitude sites showing synchronized oscillations, but with a paleoenvironmental expression or climatic signal different from the circum-North Atlantic oscillations.

2. High southern latitude sites showing oscillations diachronous from those characteristic for the circum-North Atlantic region.

3. Mid-southern latitude and tropical sites showing no oscillations or stepwise changes diachronous from the circum-North Atlantic oscillations.

In general, the late glacial climate oscillations, clearly expressed in polar and tropical ice cores (Thompson et al., 1998; Jouzel et al., 1987) and in marine cores (Bard et al., 1987, 1997; Bond et al., 1993), are rarely recorded as clearly in terrestrial records from the Americas. Instead, the character and amplitude of these oscillations in terrestrial records appear to be greatly affected by site specifics.

During the time interval from 15,500-14,500 cal. B.P., vegetation records show that temperatures were low in the high latitudes of the Northern Hemisphere, in Alaska, in the northwest Pacific, in the northeast Atlantic, and at high-elevation sites in the Colombian and Venezuelan Andes. Conditions were wet in the midlati-tudes and in tropical and subtropical forest regions, al though there is a severe lack of modern analogs for this period. Conditions were dry in southern South America and along the Venezuelan/southeast Caribbean coast.

During the B0lling-Aller0d interval (14,500-12,700 cal. B.P.), conditions became wetter in the Americas, except at sites on the Olympic Peninsula and at Lake Valencia.

The case of the next interval (12,700-11,000 cal. B.P.), which corresponds to the Younger Dryas chronozone, is different and better documented because it has been subject to more detailed analysis and even has led to major controversies (Ashworth and Markgraf, 1989; Peteet et al., 1990, 1993; Curtis and Hodell, 1993; Kuhry et al., 1993; Markgraf, 1993; Mathewes, 1993; Francou et al., 1995; Gasse et al., 1995; Hansen, 1995; Heine, 1993, 1995; Islebe et al., 1995; Leyden, 1995; Mayle and Cwynar, 1995; Osborn et al., 1995; Peteet, 1995; Thompson et al., 1995; van der Hammen and Hooghiemstra, 1995; Clapperton et al., 1997; Menounos and Reasoner, 1997; Yu and Eicher, 1998). Changes in vegetation type and composition shown during this interval that can be attributed to temperature reversals are not clearly observed in Central America (except in Costa Rica; Hooghiemstra et al., 1992), tropical lowlands, and the Southern Hemisphere midlatitudes. Moist forests are well developed in these regions, showing few signs of short-term environmental fluctuations that might correlate with the well-documented climatic cooling in the circum-North Atlantic region. On the other hand, records from the high northern Andean and Central American cordilleras show multiple stepwise vegetation changes, suggesting repeated temperature reversals. The same patterns are also recorded in ice core records from Huascarán, Peru, and Sajama, Bolivia (Thompson et al., 1995, 1998), and by evidence of past glacier fluctuations in the Andean cordilleras (Seltzer, 1994; Francou et al., 1995; Clapperton et al., 1997). Abrupt oscillations in forest development are observed in both the North American Northwest and Southwest. In southern South America, repeated high-amplitude paleoenvironmental changes are also recorded, although the timing is apparently out of phase with the circum-North Atlantic oscillations (Markgraf, 1993; Heusser, 1995; Ariztegui et al., 1997; Moreno, 1997; Markgraf and Bianchi, 1999; Denton et al., 1999; Moreno et al., 1999).

Are all the changes evidenced in the PEP 1 transect synchronous and, therefore, can they be attributed to the same climatic forcing? This unanswered question has recently received new interest. Because of the fluctuations' high frequency and their perhaps global expression (Peteet, 1995; Broecker et al., 1998; Lowell et al., 1995), events like the Younger Dryas or B0lling-Aller0d interval cannot be attributed to insolation forc ing (Milankovitch cycles). It is now generally thought that the reorganization of the ocean-atmosphere interaction in the Atlantic Ocean (changes in the intensity of the thermohaline circulation) must have played a major role in synchronizing climate oscillations during full glacial and late glacial times (Broecker et al., 1985; Broecker and Denton, 1989). It appears that synchrone-ity of late glacial oscillations is well established among Greenland, the Cariaco trench, and the Andean cordilleras (in Costa Rica, Colombia, and Peru). We suggest that these oscillations affected climates even as far equator-ward as the southern extension of the Intertropical Convergence Zone (ITCZ), with a speculative southernmost limit of the ITCZ during the Northern Hemisphere winter months (Fig. 7). The mechanism explaining the observed climate changes on land would relate to increased (or decreased during the warm

0 0

Post a comment