Elsevier

Quaternary Science Reviews

Volume 63, 1 March 2013, Pages 59-72
Quaternary Science Reviews

Astronomical tuning of long pollen records reveals the dynamic history of montane biomes and lake levels in the tropical high Andes during the Quaternary

https://doi.org/10.1016/j.quascirev.2012.11.004Get rights and content

Abstract

Long pollen records from two sediment cores of the basin of Bogotá (Colombia) are presented, reflecting the dynamic history of environmental and vegetation changes in the tropical high Andes during the Quaternary. An astronomically tuned age model has been developed by using the visual correlation of the temperature-constrained record of Alnus with the benthic δ18O stacked record of ODP Site 846 and 849 for the last 1 million year (Ma). In addition, spectral analysis of the arboreal pollen percentages (AP%) prior to 1 Ma shows two consistent peaks at 7.6 and 9.5 m, which could either be interpreted as reflecting the 19 and 23-kyr components of the precession cycle or the 41-kyr cycle of obliquity. Evidence for precession forcing comes from the sum of the filtered 7.6 and 9.5 m cycles, revealing a long-term modulation that can be linked to the short-term (100-kyr) and long-term (400-kyr) eccentricity cycles. A precession-forced scenario, however, is not consistent with zircon fission-track ages and is difficult to explain in terms of climate forcing mechanisms. On the contrary, an obliquity control of the AP% record is consistent with the fission-track ages and with glacial–interglacial-bound temperature changes within the study region. Accordingly, it appears that the Funza09 record extends back to ∼2.25 Ma during which four periods with distinct different depositional environments could be identified, following a chronological poorly constrained interval of fluvial and fluvio-lacustrine sediments. From ∼2.25 to ∼1.47 Ma sediment deposition occurred in an area of wetlands dissected by fluvial channels and swamps. The strongest subsidence of the basin occurred between ∼1.47 and ∼1.23 Ma when a proper lake developed. Lacustrine sediments accumulated in water up to 50 m deep between ∼1.23 and ∼0.86 Ma. Hereafter, water depth was generally lower than 50 m, but fluctuated in conjunction with the 100-kyr dominated glacial–interglacial variability during the middle and late Pleistocene. The evolutionary changes of the páramo and montane forest biomes are described in terms of five characteristic stages. Most of the Pleistocene vegetation has no analogue to modern assemblages principally due to the late immigration events of Alnus (1.01 Ma) and Quercus (0.43 Ma) and we conclude that forest composition similar to modern was not established until the Last Interglacial. However, modern ecological constraints of suites of taxa that formed the vegetation of the pre-MIS 5 part of the record allow a reconstruction of environmental and climate change.

Highlights

► Long continental pollen record spanning almost the entire Pleistocene. ► A 1641 points timeseries, based on published and unpublished pollen data. ► Land–sea correlation, astronomical tuning, and spectral analysis. ► Obliquity vs. precession driven vegetation dynamics on tropical mountains. ► Evolution of the Andean vegetation and basin development.

Introduction

One of the major achievements in the study of Earth system history is the reconstruction of global climate change over millions of years derived from deep sea sedimentary archives (Zachos et al., 2001; Walker and Lowe, 2007). In addition, ice archives from tropical glaciers (Thompson et al., 1998) and high-latitude ice sheets (Grootes et al., 1993; Jouzel et al., 2007; Loulergue et al., 2008) have provided a thorough picture of Pleistocene climate variability in unprecedented detail. There is a dearth, however, of designated climate records from terrestrial archives, and in particular from tropical areas, which prohibit a full understanding of the driving mechanisms behind long-term climate change and its impact on terrestrial biota. Though terrestrial pollen records were among the first to show a major step in global climate cooling about ∼2.6 million years ago (Ma) associated with the build-up and retreat of the large Northern Hemisphere ice caps (e.g. Godwin, 1956; Frenzel, 1968; Van der Hammen et al., 1971; Zagwijn, 1975), the fragmented character of continental records have hampered attempts to develop Quaternary reference sequences of terrestrial changes, partly due to the absence of robust and independent age constraints (e.g. Tzedakis et al., 2006). Land-based pollen sequences are therefore often compared with marine records on the basis of climatostratigraphic correlations (e.g. Tzedakis et al., 1997, 2001, 2004; Ruddiman, 2006; Bailey, 2009), while pollen records derived from strategically-placed marine sequences have allowed a direct marine–continental correlation on both orbital and millennial timescales, which refined our understanding of phase relationships between different components of the Earth system during the middle to late Pleistocene (e.g. Tzedakis et al., 2004; Margari et al., 2010).

Here we will elaborate on long pollen records derived from two sediment cores, Funza-1 and Funza-2, of the Bogotá basin (4°N, 2550 m altitude) in the tropical high Andes (Fig. 1). At present, the high plain of Bogotá represents the floor of a Pleistocene lake, which has been studied now for five decades, yielding a comprehensive data set of pollen and sediment fractions (Van der Hammen and González, 1960, 1963; Hooghiemstra, 1984, 1989; Hooghiemstra and Sarmiento, 1991; Van Veer and Hooghiemstra, 2000; Torres et al., 2005; Torres, 2006). The Bogotá basin consolidated after the strongest uplifting phase of the Eastern Cordillera in the Northern Andes (Van der Hammen et al., 1973; Helmens, 1990; Wijninga, 1996; Mora et al., 2008, 2010; Hoorn et al., 2010; Helmens, 2011). Although the lower part of the Funza-2 core (586–485 m) is interpreted as fluvial and may contain gaps in sedimentary record, sediments seem to have accumulated on a long-term scale without major interruptions (Hooghiemstra, 1984; Torres et al., 2005). At approximately 27 ka the basin lost its permanent lake; most plausibly the basin was overfilled with sediments giving rise to a large wetland area on the basin floor (Hooghiemstra, 1984). Much of the present-day high plain has been drained.

Dating of volcanic ash beds from previous analysis has repeatedly changed the age model of the studied sediments, because the wind-transported finest ash fractions were heavily contaminated with lacustrine deposits giving largely uncertain results (Andriessen et al., 1993). In this paper, we will focus on the cyclostratigraphic aspects, i.e. identification of the main astronomical-related frequencies in the composite pollen-based records of Funza-1 and Funza-2 to build a tuned time scale for the Quaternary sediments of Bogotá basin. Such an approach was successful in the adjacent Fúquene basin where a 60-m deep composite record reflecting the last 284,000 years (284 ka) was studied (Groot et al., 2011; Bogotá-A et al., 2011a,b; Vriend et al., 2012). Our time scale will give constraints on the time series of grain size analysis and aquatic vegetation, which will shed new light upon the development of sedimentary environments through time. In particular, we re-visit the history of the lake and show that the lake had not developed before 1.25–1.47 Ma. This led to a re-interpretation of the lower part of the pollen record. We show that the previously postulated hypothesis of an essentially different temperature-constrained altitudinal vegetation distribution is unnecessary to explain the record and we give a new environmental interpretation to the lower part of the pollen record. Finally, we will place the Pleistocene evolution of the montane forest and páramo biomes of the northern Andes into our new time framework.

Section snippets

Materials and methods

Funza-1 and Funza-2 were drilled near the village of Funza (4°50′N, 74°12′W, 2550 m.a.s.l.) located at the deepest part of the Bogotá basin (Fig. 1). Core Funza-1 was collected with rotary drilling from a mobile Portadrill truck. The site lies some 600 m from a wetland area bordering the small Bogotá River. Cores were collected in 50 cm increments and 6 cm diameter. The core reached to 357 m depth where technical problems with over-pressured ground water prevented deeper sediments from being

Age model construction

The initial age model of Funza-1 was based on 9 potassium-argon ages and 10 fission-track ages, all of poor statistical quality because only the finest wind-transported fraction of volcanic glass shards from the Central Cordillera were present in the sediments (Hooghiemstra, 1984). In addition a comparison was made between the pollen sequence of Funza-1 with a few European records to evaluate the age control of the core and to portray its climatologically importance (Hooghiemstra, 1984, 1989).

Climate evolution and time scale implications

At present it is difficult to judge whether the P-tuning or O-tuning model is to be preferred for the lower part of the Funza09 composite record, because independent age constraints are lacking besides the fission-track dates of the ash layers by Andriessen et al. (1993). At first sight, the consistency of the 7.6 m and 9.5 m cycles suggests a dominant precession forcing of the vegetation changes in the Bogotá basin prior to ∼1 Ma, which is clearly shown in the wavelet diagram of the AP% (Fig. 9

Conclusions

Sediments in the Bogotá basin have been explored for five decades and constitute a rich source of information on Pleistocene climate change and evolution of plant associations. However, different types of proxies and dating of collected volcanic ashes have been unable to provide a robust age model for sediments beyond radiocarbon time control. This paper represents a significant step forward in addressing the persistent uncertainty in the chronology of the Funza sequence. For the upper 255 m,

Acknowledgements

We acknowledge the Colombian Geological Survey (Ingeominas, Bogotá) for the drilling cores Funza-1 (1974) and Funza-2 (1988) and COLCIENCIAS (Bogotá) for financial support to the drilling operation. The Netherlands Foundation for Scientific Research (project H 75-284 to H. Hooghiemstra, project W 75-168 to H. Hooghiemstra/E.T.H. Ran) and the University of Amsterdam (projects to T. Van der Hammen/O. Hulshof and to H. Hooghiemstra/V. Torres) for supporting this long continental drilling project

References (59)

  • P.C. Tzedakis et al.

    Comparison of terrestrial and marine records of changing climate of the last 500,000 years

    Earth and Planetary Science Letters

    (1997)
  • P.C. Tzedakis et al.

    Establishing a terrestrial chronological framework as a basis for biostratigraphical comparisons

    Quaternary Science Reviews

    (2001)
  • P.C. Tzedakis et al.

    The last 1.35 million years at Tenaghi Philippon: revised chronostratigraphy and long-term vegetation trends

    Quaternary Science Reviews

    (2006)
  • T. Van der Hammen et al.

    Palynological record of the upheaval of the Northern Andes: a study of the Pliocene and Lower Quaternary of the Colombian Eastern Cordillera and the early evolution of its high-Andean biota

    Review of Palaeobotany and Palynology

    (1973)
  • R.J. Bailey

    Cyclostratigraphic reasoning and orbital time calibration

    Terra Nova

    (2009)
  • H.J.B. Birks et al.

    Quaternary Palaeoecology

    (1980)
  • Bogotá-A, R.G., Hooghiemstra, H., Berrio, J.C., 2011b. An ultra-high resolution multi-proxy record from lake Fúquene...
  • A.M. Cleef

    The phytogeographical position of the neotropical vascular paramo flora with special reference to the Colombian Cordilera Oriental

  • A.M. Cleef et al.

    The savanna relationship in the Andean paramo flora

    Opera Botanica

    (1993)
  • B. Frenzel

    Grundzüge der Pleistozänen Vegetationsgschichte Nord-Eurasiens

    (1968)
  • H. Godwin

    The History of the British Flora

    (1956)
  • Grabandt, R., 1985. Pollen rain in relation to paramo vegetation in the Colombian Cordillera Oriental. Ph.D. thesis,...
  • M.H.M. Groot et al.

    Ultra-high resolution pollen record from the northern Andes reveals rapid shifts in montane climates within the last two glacial cycles

    Climates of the Past

    (2011)
  • P.M. Grootes et al.

    Comparison of oxygen isotope records from the GISP2 and GRIP Greenland ice cores

    Nature

    (1993)
  • K.F. Helmens

    Neogene-Quaternary geology of the high plain of Bogotá, Eastern Cordillera, Colombia (stratigraphy, palaeoenvironments and landscape evolution)

    Dissertationes Botanicae

    (1990)
  • H. Hooghiemstra

    Vegetational and climatic history of the high plain of Bogotá, Colombia

    Dissertationes Botanicae

    (1984)
  • H. Hooghiemstra

    Immigration of oak into northern South America: a paleo-ecological perspective

  • H. Hooghiemstra et al.

    Present vegetation of the area of the high plain of Bogotá

  • H. Hooghiemstra et al.

    Long continental pollen record from a tropical intermontane basin: Late Pliocene and Pleistocene history from a 540-meter core

    Episodes

    (1991)
  • Cited by (67)

    • A 1.8 million year history of Amazon vegetation

      2023, Quaternary Science Reviews
      Citation Excerpt :

      The consistent occurrence of Alnus combined with the high number of up to 10 pollen grains per sample in the CDH-79 record (Fig. 4), strongly supports our interpretation that it reached the Amazon Basin in the earliest Pleistocene, agreeing with extra-Amazonian records from northernmost South American records (e.g. Germeraad et al., 1968; Pocknall et al., 2001). The presence of Alnus as early as 1.8 Ma at CDH-79 stands in contrast to its first well-dated occurrence at 1 Ma in the Bogotá Basin of Colombia at an elevation of 2550 m (Hooghiemstra and Cleef, 1995; van der Hammen and Hooghiemstra, 2000; Torres et al., 2013). Previous studies indicate that Alnus migrated from North America to southern Mexico during cooling climates of the Pliocene (Graham, 1976, 1999), but other records in Central America remain unconfirmed (e.g. Neogene of Panama; Graham, 1999).

    • Alternate Atlantic forest and climate phases during the early Pleistocene 41 kyr cycles in southeastern Brazil

      2022, Quaternary Science Reviews
      Citation Excerpt :

      The basin was composed of wetlands and swamps until ca.1.3 Myr when faster subsidence of the basin floor transformed the zone into a big deep lake that lasted until the MPT (Torres et al., 2013). Upward shifts in the tree line in the region were paced by the succession of the glacial interglacial cycles (Torres et al., 2013). Despite the existence of such record, there is no clear description of the environmental dynamics and the climatic system that dominated during the early Pleistocene in South America.

    View all citing articles on Scopus
    View full text