Elsevier

Quaternary Science Reviews

Volume 122, 15 August 2015, Pages 51-62
Quaternary Science Reviews

Origin and dynamics of the northern South American coastal savanna belt during the Holocene – the role of climate, sea-level, fire and humans

https://doi.org/10.1016/j.quascirev.2015.05.009Get rights and content

Highlights

  • A 11,500 yr old record from northern South American coastal savanna belt.

  • Role of climate, sea level, fire and human on the coastal savanna belt.

  • Evidence of a mono-dominant forest without modern analog at the onset of Holocene.

  • New insights into the early Holocene sea level rise near the mouth of Amazon river.

  • Three phases of marked environmental changes due to shift in atmospheric convection.

Abstract

Presence of a coastal savanna belt expanding from British Guiana to northeastern Brazil cannot be explained by present-day climate. Using pollen and charcoal analyses on an 11.6 k old sediment core from a coastal depression in the savanna belt near the mouth of the Amazon River we investigated the paleoenvironmental history to shed light on this question. Results indicate that small areas of savanna accompanied by a forest type composed primarily by the genus Micropholis (Sapotaceae) that has no modern analog existed at the beginning of the Holocene. After 11,200 cal yr BP, savanna accompanied by few trees replaced the forest. In depressions swamp forest developed and by ca 10,000 cal yr BP replaced by Mauritia swamps. Between 8500 and 5600 cal yr BP gallery forest (composed mainly of Euphorbiaceae) and swamp forest succeeded the treeless savanna. The modern vegetation with alternating gallery forest and savanna developed after 5600 cal yr BP. We suggest that the early Holocene no-analog forest is a relict of previously more extensive forest under cooler and moister Lateglacial conditions. The early Holocene savanna expansion indicates a drier phase probably related to the shift of the Intertropical Convergence Zone (ITCZ) towards its northernmost position. The mid-Holocene forest expansion is probably a result of the combined influence of equatorwards shift of ITCZ joining the South Atlantic Convergence Zone (SACZ). The ecosystem variability during the last 5600 cal yr BP, formed perhaps under influence of intensified ENSO condition. High charcoal concentrations, especially during the early Holocene, indicate that natural and/or anthropogenic fires may have maintained the savanna. However, our results propose that climate change is the main driving factor for the formation of the coastal savanna in this region. Our results also show that the early Holocene sea level rise established mangroves near the study site until 7500 cal yr BP and promoted swamp formation in depressions, but did not influence the savanna vegetation.

Introduction

Present-day savannas occupy approximately 16.1 million km2, or 11.5% of the global landmass. The proportion of trees and grasses in savannas is related to climate and land use practice. In addition, extensive biomass burning during the dry season plays an important role to inhibit tree growth (Scholes and Hall, 1996, Murphy and Bowman, 2012). Considering the large difference between above ground carbon storage capacity of treeless grasslands (2 tons C/ha) and woodland savannas (30 tons C/ha), besides the huge amount of CO2 that may be emitted to atmosphere by biomass burning of savannas (estimated at 0.5–4.2 Gt C per year globally) (Grace et al., 2006), plans for CO2 management need to consider climatic/anthropogenic influences which trigger change from savanna with arboreal taxa (cerradão) to treeless grasslands (campo limpo).

A narrow strip of savanna known as “coastal savanna belt” in northern South America is found along the coast of British Guiana, Surinam, French Guiana, and in State of Amapá, Marajo Island and in part in the state of Pará. This discontinuous belt about 2000 km long is disrupted locally by other types of coastal vegetation (Fig. 1a). Based on the meteorological data (NOAA), annual precipitation in the coastal savanna belt is in a range similar to most of Amazon regions (between 1750 and 3500 mm) (Snow, 1976, Nimer, 1989, Weischet, 1996). Therefore instead of savanna presence of Amazon rainforest would be expected. Because of its considerable area the savanna's existence and its dynamics should have a substantial effect on the regional carbon budget.

Several palaeo-environmental studies using pollen and charcoal analyses have previously been carried out to investigate the history of the coastal area in northern South America. Although these studies are mainly focused on mangrove development, useful clues can be obtained regarding savanna/forest/mangrove interaction. Records from Guyana (Van der Hammen, 1963), Suriname (Wijmstra, 1971) and French Guiana (Tissot and Marius, 1992) show expansion of savanna during the full glacial period (and shoreline regression) and mangrove development during the interglacial period (and sea level transgression). Swamp savanna with dominance of Poaceae and Cyperaceae was present during the last c 5700 cal yr BP in the coastal regions of Guyana, Suriname, and French Guiana (Behling and Hooghiemstra, 2001). In Amapá State lacustrine littoral records from lakes Tapera and Marcio (Toledo and Bush, 2007) also demonstrate changes in vegetation from closed forests with swamp taxa to open flooded savanna at c 4750 cal yr BP. In the pollen record of Lago Arari, on Marajó Island in the mouth of the Amazon River in northeastern Pará there is a marked change from the more or less closed to open swamp savanna and forest at ca 7400 cal yr BP (Absy, 1985). Another study from Lake Arari investigated four sediment cores, which reveal replacement of mangrove by herbaceous vegetation at 2300 cal yr BP and an expansion of herbs during the last 1000 years (Smith et al., 2012). In the Southern Hemispheric part of the coastal area, in Lago Crispim (Behling and Costa, 2001), Lagoa da Curuca (Behling, 2001) and Lagoa do Caco (Ledru et al., 2001, Pessenda et al., 2005) Holocene started with arboreal taxa dominating in the vegetation, which with different timing (due to different latitudinal position) gradually were accompanied by swamp trees and finally replaced by open vegetation.

In addition to being interesting for palaeoenvironmental research, lands on the eastern Amazonia on the Amazon River bank or near the river mouth host many archeological surveys. Late Pleistocene Paleo-Indian camp side at Monte Alegre in the eastern Brazilian Amazon documents presence of ecologically adapted foragers with presumably limited big-game hunting habit (Roosevelt et al., 1996). In another study on Marajo Island, Roosevelt and her colleagues (1991) found that mound builders have occupied the alluvial floodplains of the Lower Amazon from A.D. 400–1300.

This work presents the analysis of a 750 cm-long sediment core taken from a key area of the coastal savanna belt near the mouth of the Amazon River, which forms an ecotone between Amazon rainforest and coastal vegetation. In order to test different hypotheses concerning probable driving forces for the development of a savanna belt, such as sea level change, climate, fire, human and edaphic factors, this continuous pollen and charcoal record was analyzed and compared with other results from northern South America.

Section snippets

Study area

The core named Curiau (CUR), was collected from a small Mauritia swamp 15 m in diameter (00°12′30.3″ N, 51°01′12.1″ W, 5 m a.s.l) located 16 km north of Macapá City in the south of Amapá State near the mouth of Amazon river (Fig. 1b). The studied swamp is surrounded by small hills which have an elevation up to 22 m a.s.l.

Material and methods

The core CUR was taken from a Mauritia swamp in November 2004 using a Russian Corer and was transported to the Federal University of Pará and stored in a refrigerator after subsampling.

Stratigraphy and chronology

In the sediment core CUR two major types of deposits are recognizable: The bottom of the core (750–700 cm) contains gray sandy sediments with small amounts of organic material. Between 700 and 0 cm the sediment is composed of dark gray material with a high organic content. In the uppermost 30 cm some bioturbation may occur. In general, from the base to the top of the core accumulation rate decreased. However, between 700 and 550 cm accumulation rate is three-fold and shows a strong change in

Beginning of the Holocene (Zone CUR-I, 11,500–11,200 cal yr BP)

The pollen assemblages of this period indicate a landscape occupied by a mono-dominant arboreal community composed of Micropholis and mixed with sparse trees of Protium and Myrtaceae family, which probably grew in the depressions. Small patches of savanna probably covered the top of the hills along the Pleistocene plains. Today Micropholis is found mainly both in the lowland and upland rainforest of South America (Lorenzi, 1992, Lorenzi, 1998). Therefore it can be concluded that the formation

Discussion

The data reveal that during the last 11,500 years, savanna was an integral part of the landscape in the Amapá coastal region. However, it expanded after the extinction of the specific arboreal vegetation, which covered the area at the beginning of the Holocene (>11,500–11,200 cal yr BP). During the early Holocene (11,200–8500 cal yr BP) savanna was dominant and occurred on the elevated areas of the hills along Pleistocene plain and forest swamp established along palaeo-channels. At

Conclusion

The following main conclusions can be made from this study:

  • 1

    Similar to many regions in the lowland Amazonia, a unique forest type without modern analog grew near the mouth of Amazon River at the beginning of the Holocene.

  • 2

    Savanna had been continuously present in the coastal area of Amapá State at least since the beginning of the Holocene (11,500 cal yr BP).

  • 3

    In this region savanna has an intermediate position between interhemispheric shifts of atmospheric convection systems, because the early

Acknowledgment

We would like to thank Dr Thomas Giesecke (University of Göttingen), who kindly supervised statistical analysis of this investigation. We sincerely appreciate Professor Vera Markgarf (INSTAAR, University of Colorado Boulder) for her sage advice on both language and scientific aspect of the paper. We thank also Suzette G.A. Flantua manager of Latin American Pollen Database from the Institute for Biodiversity and Ecosystem Dynamics (IBED), University of Amsterdam, for the useful advice on the

References (77)

  • E.C. Grimm

    CONISS: a Fortran 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares

    Comput. Geosci.

    (1987)
  • M.P. Ledru et al.

    Late-glacial cooling in Amazonia inferred from pollen at Lagoa do Caçó, northern Brazil

    Quat. Res.

    (2001)
  • R. Marchant et al.

    Distribution and ecology of parent taxa of pollen lodged within the Latin American pollen Database

    Rev. Palaeobot. Palyno.

    (2002)
  • L.S. Moreira et al.

    Holocene paleoenvironmental reconstruction in the eastern Amazonian Basin: comprido Lake

    J. South Amer. Earth Sci.

    (2013)
  • C.B. Smith et al.

    Holocenic proxies of sedimentary organic matter and the evolution of Lake Arari-Northern Brazil

    Catena

    (2012)
  • H. Wanner et al.

    Quat. Sci. Rev.

    (2008)
  • M.L. Absy

    Palynology of Amazonia: The history of the forests as revealed by the palynological record

  • H. Behling et al.

    Holocene Amazon rainforest–savanna dynamics and climatic implications: high-resolution pollen record from Laguna Loma Linda in eastern Colombia

    J. Quat. Sci.

    (2000)
  • H. Behling

    First report on new evidence for the occurrence of Podocarpus and possible human presence at the mouth of the Amazon during the Late-glacial

    Veget Hist. Archaeobot

    (1996)
  • H. Behling

    Late Quaternary environmental changes in the Lagoa da Curuca region (eastern Amazonia, Brazil) and evidence of Podocarpus in the Amazon lowland

    Veget Hist. Archaeobot.

    (2001)
  • J.C. Berrio et al.

    Late-Quaternary savanna history of the Colombian Llanos Orientales from Lagunas Chenevo and Mozambique: a transect synthesis

    Holocene

    (2002)
  • J.P. Bradbury et al.

    Late Quaternary environmental history of Lake Valencia, Venezuela

    Science

    (1981)
  • L.M.M. Carreira et al.
  • F.P. Carvalho et al.

    Atla Zoneamento Costeir Estuarino Do Estado Do Amapá. Macapá, PNMA/SQA/MMA

    (2006)
  • J.W.A. Castro et al.

    Sea-level fluctuations and coastal evolution in the state of Rio de Janeiro, southeastern Brazil

    An. Acad. Bras. Ciênc.

    (2014)
  • P.A. Colinvaux et al.

    A long pollen record from lowland Amazonia

    For. Cool. Glacial Times. Sci.

    (1996)
  • P.A. Colinvaux et al.

    Amazon Pollen Manual and Atlas – Manual e Atlas Palinológico da Amazônia

    (1999)
  • S.V. Costa Neto

    Relatório de vegetação: Subsídio ao diagnostico sócio ambiental. Relatório Técnico. Macapá, IEPA/GERCO

    (2004)
  • S.V. Costa Neto et al.

    Vegetação do setor costeiro estuarino do estado do Amapá. Instituto de Pesquisas Científicas e Tecnológicas do Estado do Amapá. Governo do Estado do Amapá. Cap. 5. Projeto Zoneamento Econômico-Ecológico do setor costeiro estuarino: Diag-nóstico sócio ambiental participativo do setor costeiro estuarino

    (2004)
  • S.V. Costa Neto et al.

    Macrófitas aquáticas das Regiões dos Lagos do Amapá, Brasil

    Rev. Bras. Biociênc.

    (2007)
  • J.H. Curtis et al.

    Climate change in the Lake Valencia Basin, Venezuela, ∼12,600 yr BP to present

    Holocene

    (1999)
  • E. Delhaize et al.

    Aluminum toxicity and tolerance in plants

    Plant Physiol.

    (1995)
  • Dias et al.

    Mid-Holocene climate of tropical South America: a model-data approach

  • T.H. Donders et al.

    Mid- to late-Holocene El Niño-Southern oscillation dynamics reflected in the subtropical terrestrial realm

    PNAS

    (2005)
  • K. Faegri et al.

    Textbook of Pollen Analyses

    (1989)
  • Y. Fan et al.

    Global patterns of groundwater table depth

    Sci

    (2013)
  • R.G. Faribanks

    A 17,000-years glacio-eustatic sea level record: influence of glacial melting rates on the Younger Dryas event and deep-ocean circulation

    Nature

    (1989)
  • J. Grace et al.

    Productivity and carbon fluxes of tropical savannas

    J. Biogeogr.

    (2006)
  • Cited by (0)

    View full text