This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Contrasting responses of two Caribbean mangroves to sea-level rise in the GuajiraPeninsula (Colombian Caribbean)

Ligia E. Urrego a,⁎, Alexander Correa-Metrio b, Catalina González c, Ana R. Castaño d, Yusuke Yokoyama e

a Departamento de Ciencias Forestales, Universidad Nacional de Colombia, Sede Medellín A.A. 568 Medellín, Colombiab Instituto de Geología, Universidad Nacional Autónoma de México, México City DF 04510, Méxicoc Departamento de Ciencias Biológicas, Universidad de los Andes, A.A. 4976 Bogotá, Colombiad ICES (International Center For Earth Sciences), Mendoza. Argentinae Department of Earth and Planetary Sciences, University of Tokyo, Japan

a b s t r a c ta r t i c l e i n f o

Article history:Received 28 June 2012Received in revised form 27 November 2012Accepted 29 November 2012Available online 7 December 2012

Keywords:MangrovesPollen analysisSea-levelCoastal progradationCoastal erosionColombian Caribbean

Local geomorphology plays an important role in the response of mangrove communities to both sea-level riseand precipitation changes, mostly because it exerts an important control of the erosion–progradation bal-ance. Here we present contrasting changes in the palynological records of two sediment cores retrievedfrom locations occupied by mangroves in the northeastern Caribbean coast of Colombia: Calancala andNavío Quebrado. Sediments were treated with standard procedures used in palynologycal analysis, anddated with AMS 14C and 210Pb. Age-depth models of both cores were constructed by linear interpolation be-tween dated samples. The main modern differences between the studied sites are fresh water availability andsediment supply throughout the year. While Calancala, located at the Ranchería River Delta, has direct freshwater influence throughout the year, coastal lagoon Navío Quebrado is only sporadically influenced by freshwater sources. According to our findings, two main factors contributed to mangrove establishment in theColombian Caribbean: i) high and accelerated sea-level rise around 6000 cal yr BP originating coastallagoons that favored mangrove establishment, and ii) marine still stand and high precipitation around3000 cal yr BP leading to a second wave of mangrove expansion. A drought between about 2850 and2450 cal yr BP and the strengthening of easterly trade winds caused a change from Rhizophora-dominatedto Avicennia-dominated stands. Increased precipitation after 2500 cal yr BP and subsequent increase of fluvi-al sediment input promoted recovery of Rhizophora-dominated mangroves. Sea-level rise during the last150 yr has caused landward mangrove expansions in deltas and progradation of coastal environments.However, the general pattern shows a net loss of mangrove vegetation as the result of coastal erosion.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

Mangroves are ecosystems distributed along tropical and subtropicalcoasts worldwide (Giri et al., 2011). They are typically composed ofplants physiologically and morphologically adapted to living in theintertidal zone, which is periodically inundated by brackish water.Mangrove communities develop in a diverse range of sedimentary envi-ronments, ranging from highly depositional alluvial habitats to oceanicislands with little or no permanent contribution of allochtonous sedi-ments. Their establishment and growth are determined bymany factors,the most important being high air temperatures, tidal and fresh waterflooding, salinity, geomorphological setting, soil type, nutrient andlight availability, and carbon dioxide (CO2) concentrations (Krauss et

al., 2008). Vegetation structure and composition of mangroves resultfrom the differential response of both mangrove and associated speciesto the environmental setting, although regional and local flora diversityas well as disturbance regimes may also play important roles (Lugo andSnedaker, 1974; Urrego et al., 2009a).

Given their ubiquity mediated by particular environmentalconditions, mangroves are sensitive to environmental changes. Modernrelationships between mangrove species and the environment are as-sumed to have operated similarly during the Quaternary, making pos-sible the inference of past coastal environments through the analysis offossil vegetation (Hooghiemstra et al., 2006; Ellison, 2008). Of specialinterest is the reconstruction of changes in sea-level and/or coastalprogradation or erosion (Hooghiemstra et al., 2006; Ellison, 2008) atvarious time scales. In fact, because of the specialized distribution ofparticular mangrove species along environmental gradients, they canbe used as proxies for specific ecological conditions at local scales(Tomlinson, 1995; Urrego et al., 2009a). In palynological reconstruc-tions from the Caribbean, mangrove forests are typically representedby four true tree mangrove species which are usually found in the

Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

⁎ Corresponding author. Tel.: +57 4 4309133; fax: +57 4 4309018.E-mail addresses: leurrego@unal.edu.co (L.E. Urrego), acorrea@geologia.unam.mx

(A. Correa-Metrio), c.gonzalez2579@uniandes.edu.co (C. González),anarosacastanoganan@icesmalargue.org (A.R. Castaño), yokoyama@ori.u-tokyo.ac.jp(Y. Yokoyama).

0031-0182/$ – see front matter © 2012 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.palaeo.2012.11.023

Contents lists available at SciVerse ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r .com/ locate /pa laeo

Author's personal copy

following order of relative importance: Rhizophora mangle, Avicenniagerminans, Laguncularia racemosa and Conocarpus erectus (Urrego etal., 2010). However, other non-arboreal species are also importantcolonizers at the beginning of the forest succession and/or indica-tors of degraded mangroves. Such is the case of Batis maritima,Sesuvium portulacastrum and the “mangrove fern” Acrostichum aureum(Tomlinson, 1995). Although associated species are usually poorlyrepresented, they are important proxies for specific ecological featuresand successional stages.

High percentages of Rhizophora pollen in sediment samples usuallyindicate either the proximity of pure mangrove stands (Muller, 1959;Behling et al., 2001), or high flooding levels and shorter distances tothe sea (Urrego et al., 2010). Conversely, low Rhizophora proportions in-dicate tidal or fluvial transport of pollen grains, and/or mangroves locat-ed at foothills close to the transition to hinterland or highly disturbedmangrove forests (Urrego et al., 2009a, 2010). Contrastingly, Avicenniapollen, though usually underrepresented in pollen records, typically re-flects saline environments (Urrego et al., 2010), hurricane-disturbedmangroves or severe droughts and plagues (González et al., 2010).High percentages of Laguncularia and Acrostichum indicate disturbedopen canopies and early successional stages following mangrove distur-bance (Medina et al., 1990; Tomlinson, 1995). Also, the presence of her-baceous and salt marsh pollen species in the Caribbean usually indicatesalinity gradients (González and Dupont, 2009a), early successionalstages after disturbance (González et al., 2010), or marine transgressiveevents (González and Dupont, 2009a; Urrego et al., 2009b). A specialfeature in coastal plains is the sequence of Amaranthaceae (formerlyChenopodiaceae, sensu Stevens, 2001) – Poaceae – Cyperaceae, whichindicates salt marsh expansion and sediment relocation under directtidal influence usually followed by mangrove forest development(González and Dupont, 2009b).

Mangrove pollen records have provided indirect reconstructions ofsea-level and/or coastal progradation–erosion balance (Hooghiemstraet al., 2006; Ellison, 2008), and today most of our knowledge onHolocene sea-level variability in the Caribbean derives from past vegeta-tion reconstructions on coastal environments (e.g. Rull et al., 1999;Ramcharan, 2004). However, in some instances, the importance of fac-tors such as physiographic settings and precipitation has overwhelmedsea-level change signals (Eslami-Andargoli et al., 2009), calling for acareful interpretation of mangrove pollen records. When comparingthe existing evidence on mangrove dynamics in the Caribbean, the lackof synchronicity among mangrove establishment and sea-level changesduring the Holocene is noteworthy, highlighting the importance of in-cluding local processes in the interpretation. Based on twonewsedimen-tary records from the Colombian Caribbean, we reconstruct mangrovedynamics in the study area from the mid-Holocene to present. Also, weaddress two basic questions: i) whatwas the relationship betweenman-grove dynamics and sea-level changes? and ii)were the changes inman-grove communities from the Colombian Caribbean the result of regional(Caribbean-wide) scale processes?

2. Study area

Calancala (11° 34′30″ N, 72° 52′36″W) and Navío Quebrado (11°24′ 39″ N, 73°06′00″ W) are both located 1 m asl, in the ColombianCaribbean coast, which is a 1760-km-long coast line that extendsfrom the Gulf of Urabá to the Gulf of Venezuela, the northernmostpart of South America. Warm climates predominate in the area witha marked precipitation gradient that goes from wet to arid conditionsin the southwestern Caribbean and the Guajira Peninsula, respectively(Fig. 1). Mean annual temperature in La Guajira area is 27 °C with an-nual extremes of 15 and 38 °C (Sánchez-Páez et al., 1997). As in mostof northern South America, the Intertropical Convergence Zone (ITCZ)modulates the bimodal distribution of precipitation and consequentlythe fluvial sediment supply to the study locations. Mean annual precip-itation is 550 mm (HIMAT, 1994), with dry periods between December

and April, and July and August. The Caribbean Low Level Jet blows par-allel to the Guajira coast, inducing coastal upwelling (Andrade andBarton, 2005) and diverting rainfall from the sea.

With a 4296-km2 basin, the Ranchería River flows northwestwarddown the Sierra Nevada de Santa Marta (ca. 3000 m asl) and reachesthe coastal plain near the city of Riohacha, where the Calancala studysite is located. Downslope (between 100 and 250 m asl), the RancheríaRiver valley has a predominant southwestern-northeastern orientation.The study site is dominated by alluvial sediments that are underlain byNeogene folded sediments with some local input of aeolian sediments.The valley is enclosed by foothills and a mountainous landscape. Thecoastal sedimentary plain is slightly undulated and is mainly fed byalluvial sediments from the Cesar and Ranchería rivers, and additionalinfluence of marine and aeolian sediments. An estuary is formed bythe river mouth during the rainy season and remains closed duringthe rest of the year (IGAC, 2009). The second study site, Navío Quebradolagoon, is located 22 km south of Riohacha and is isolated from theCaribbean Sea by sand spits. It was formed by the abrasive action ofwaves, and refilled by alluvial sediments. Several intermittent creeksthat form deltas during the rainy season feed the lagoon (Fig. 1). Highcoastal erosion is driven by wave action and east-northeasterly windsthat hit the coast (Raasveldt and Tomic, 1958).

Mangrove forests in the region are under severe stress, mainlyexerted by low precipitation, non-permanent input of freshwater, andincreased high sediment input when trade winds intensify. Additional-ly, forest regeneration is threatened by human occupation through cat-tle feeding upon mangrove propagules (Lema et al., 2003), timberexploitation (Guerrero et al., 2005), extraction of products for subsis-tence of fishermen, and pollution from urban sewage and solid waste.While R. mangle proliferates on the edge of the channels and the rivermouths, mangrove communities found along the coasts are dominatedby A. germinans, with scattered L. racemosa trees (IGAC, 2009). In openareas the fern A. aureum is commonly found. Some C. erectus shrubs ap-pear at the outer boundaries of the forest, with abundant patches ofB. maritima and S. portulacastrum (Vásquez, 2000). In the seasonallyflooded zone that surrounds the mangrove forests, Eleocharisacutangula (Cyperaceae) and Leptochloa cf. uninervia (Poaceae) areabundant. Sandy beaches are dominated byHibiscus tiliaceus, Hippomeamancinella and the spiny shrub Caesalpinia coriaria (Sánchez-Páez etal., 1997) nevertheless many native species are not present todaymostly because of human disturbance.

3. Methods

Two sediment cores were raised in the Guajira region (Fig. 1)using a Dachnovsky corer. A 375-cm-long core was retrieved fromCalancala, a branch of the Ranchería River delta located near the cityof Riohacha. A second 400-cm-long core was recovered from NavíoQuebrado coastal lagoon, 22 km south east from Riohacha city. Sedi-ment samples of 1 cc were taken every 5 cm and treated with stan-dard procedures to recover palynomorphs that were subsequentlyanalyzed using transmitted-light microscopy (Erdtman, 1960; Faegriand Iversen, 1989). We aimed at a pollen sum of 250 including alltrees, herbaceous taxa from salt marshes, and the mangrove fern A.aureum (all other spores were counted but excluded from the pollensum). Although it is unusual to include fern spores in the pollen sum,we did so in this case because Acrostichum is an important componentof mangrove ecosystems and its occurrence and abundance are highlyindicative of environmental conditions, ecological processes, and suc-cessional stages. Oxidizing conditions are common under extremelydry environments such as the site where our study was conducted,causing poor pollen preservation. Thus, pollen sums of some sampleswere lower than 200 grains and should be cautiously interpreted. Infact, pollen sums of only between 100 and 150 grains were reachedfor 19% and 5% of samples of Navio Quebrado and Calancala, respec-tively (Figs. 2 and 3). Pollen sums for 21% of samples of Navio

93L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

Quebrado were between 150 and 200, and all other counts includingthose of Calancala were above 200 (29% and 86% of samples withcounts higher than 250 grains for Navio Quebrado and Calancala, re-spectively). However, relatively low plant diversity at these marineenvironments can be well represented even with low pollen sums(Hooghiemstra et al., 2006).

The chronology of Calancala core was based on 12 210Pb samplescovering the upper 144 cm, and two Accelerator Mass Spectrometry(AMS) radiocarbon dates at 155 and 199 cm depths. Chronologicalcontrol of Navío Quebrado core was established through two AMSradiocarbon dates on plant debris at 195 and 365 cm of depth.Age-depth models of both cores were constructed by linear interpola-tion between dated samples. For simplification, ages are given in cal-ibrated years Before Present (cal yr BP) for both ancient (Holocene)and recent time scales (between 1800 A.D. and present time). Zona-tion of pollen sequences was established based on stratigraphicallyconstrained clusters taking all taxa into account for calculating Eu-clidean distances. Additionally, Detrended Correspondence Analysis(DCA) was performed on both data sets to compare vegetation turn-over with the scores of the first Axis. These analyses were performedusing STATGRAPHICS (Centurion XV, Version15.1.02) and CANOCO4.5 (Ter Braak and Smilauer, 2002).

4. Results

4.1. Stratigraphy and chronology

Sediments from cores Navío Quebrado and Calancala (Table 1,Figs. 2 and 3) were highly variable in terms of color and texture.

The upper half of Navío Quebrado core consisted of finely grainedsediments, mostly clays, interspersed with shells (Fig. 2). Contrast-ingly, organic remains and sandy sediments dominated the lowerhalf of the core. According to the age model (Fig. 4), the depositionof this core started around 6300 cal yr BP (Table 2). Estimated sedi-mentation rates were rather low (between 0.05 and 0.08 cm/yr).The upper part of the Calancala core was dominated by clayey silts,whereas the lower half of the core was characterized by coarse silts(Fig. 3). According to the developed chronology, Calancala core com-prises the period from 1250 cal yr BP to present (Table 2, Fig. 4). Es-timated sedimentation rates were about 0.29 cm/yr for the periodbefore 70 cal yr BP, rising to 1.45 cm/yr from this point onwards.

4.2. Navío Quebrado pollen record

Three pollen zones were recognized for Navío Quebrado core(Figs. 2 and 4). Zone NQ I (365–255 cm, 6280–3840 cal yr BP) wascharacterized by dominance of salt marsh vegetation, mainlyCyperaceae, Poaceae and Amaranthaceae (formerly Chenopodiaceaeaccording to Stevens, 2001). The first part (Subzone NQ IA 365–340 cm, 6280–5730 cal yr BP) was characterized by high percentagesof salt marsh vegetation, mainly Cyperaceae and Poaceae, and lowproportions of mangrove vegetation (Acrostichum 10% and Rhizophorab7%). Through subzone NQ IB (340–315 cm, 5730–5170 cal yr BP),salt marsh vegetation nearly disappeared, but Amaranthaceae (12%)and mangrove taxa increased (Rhizophora 24%, Avicennia 7%,Laguncularia 3%, and Batis 5%). A third subzone NQ IC (315–255 cm,5170–3840 cal yr BP), was characterized by decreases in arborealmangroves and increases of the mangrove fern Acrostichum (25%).

Fig. 1. Study area. Study sites along the Colombian Caribbean Coast (BH: Bahía Honda, N:Navío, F:Flotante, NQ:Navío Quebrado, C: Calancala). Insets with location in Northeasterncoast and Guajira Peninsula of Navío Quebrado (Map A) and Calancala (Map B) cores.

94 L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

Fig.

2.Po

llendiag

ram

ofselected

taxa

from

coastalv

egetationfrom

Nav

íoQue

brad

oco

re,C

olom

bian

Caribb

ean.

Percen

tage

sof

taxa

arepresen

tedag

ains

tde

pth(cm)an

destimated

age(cal

yrBP

).Su

mmarypo

llendiag

ramsfrom

othe

rve

getation

arepresen

tedto

therigh

t.Sa

mples

withpo

llensu

mslower

than

150marke

dwithasterisk.

95L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

Fig.

3.Po

llendiag

ram

ofselected

taxa

from

coastalv

egetationfrom

Calanc

aco

re,C

olom

bian

Caribb

ean.

Percen

tage

sof

taxa

arepresen

tedag

ains

tde

pth(cm)an

destimated

age(cal

yrBP

).Su

mmarypo

llendiag

ramsfrom

othe

rve

getation

arepresen

tedto

therigh

t.Sa

mples

withpo

llensu

mslower

than

150marke

dwithasterisk.

96 L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

Additionally, subzone NQ IC showed decreases of Cyperaceae,antiphasing increases of Poaceae.

Zone NQ II (255–100 cm, 3840–1270 cal yr BP) was characterized bythe highest percentages of mangrove taxa, and the lowest of salt marshvegetation. Although in subzone NQIIA (255–210 cm-transition, 3840–2850 cal yr BP) both groups kept antiphasing, through subzone NQIIB(210–190 cm, 2850–2450 cal yr BP) mangrove floristic compositionchangeswere evidenced by increases of Avicennia (19%) and Cyperaceaae(26%), and decreases of other mangrove taxa. Subzone NQIIC (190–100 cm, 2450–1270 cal yr BP) was marked by a peak of Rhizophora(66%), which then decreased towards the end of the pollen zone.

Zone NQIII (100–15 cm, 1270–150 cal yr BP) was characterizedby a reduction of mangrove vegetation. While Rhizophora, Avicenniaand Araceae were not represented, Batis and Laguncularia occurredin low percentages (b3%). Strong increments of salt marsh vegeta-tion, especially Cyperaceae (36%), Poaceae (21%) and Amaranthaceae(15%), were also noteworthy.

4.3. Calancala pollen record

In the Calancala record three pollen zones were recognized, most-ly based on the changes displayed by mangrove taxa (Figs. 3 and 4).

Zone CALI (390–285 cm, 1240–720 cal yr BP) was characterized byvery low mangrove taxa percentages (b2%), and high percentages ofsalt marsh and beach vegetation (14–44%), especially Cyperaceae,Poaceae and Amaranthaceae. Sedimentation rates along this zonewere around 0.29 cm/yr. Zone CALII (285–145 cm, 720–70 cal yrBP) was characterized by low percentages of mangrove taxa(Rhizophora, Laguncularia and Batis show percentages b2%), exceptat 155 cm where Rhizophora (16%), Laguncularia (6%) and Batis (9%)showed high percentages. In this zone, sedimentation rates increasedto 0.31 cm/yr. Zone CALIII (145–15 cm, 70–50cal yr BP) showed highpercentages of mangrove taxa (Rhizophora 41%, Avicennia 6%,Laguncularia 16%, and Batis 14%), and sedimentation rates reachedthe highest values along the core (0.98 cm/yr).

5. Discussion

5.1. Lagoon formation and mangrove colonization in the northeasternColombian Caribbean

The pollen record of Navío Quebrado core shows the formationprocess of the lagoon, as well as environmental changes that tookplace in the area from 6280 cal yr BP to present. Between 6280 and3840 cal yr BP a salt marsh environment fringed by sparse mangroveforests expanded, as evidenced by high percentages of Poaceae andCyperaceae and low proportions of mangrove pollen. This sequence

Table 1Sedimentological descriptions of Navío Quebrado and Calancala Cores.

Depth(cm)

Description

Navío Quebrado15–25 Light gray silty clay with spared sea shells25–100 Light gray silty clay spotted with ironed oxid100–165 Dark grayish brown silty clay with sea shells165–200 Dark gray clayish sand200–325 Dark gray clay interspersed with dark brown bands of organic tissues325–400 Light brown clayish sand

Calancala Core0–141 Dark brown organic mud, plenty of roots and organic remains141–170 Dark brown clayish silt without organic remains170–220 Dark brown clayish silt without organic remains220–282 Grayish brown sandy mud282–293 Grayish brown sand293–361 Dark brown clayish sand361–375 Grayish brown sand

Fig. 4. Age model of cores: Navío Quebrado (A) and Calancala (B).

Table 214C and 210Pb dates for the Navio Quebrado and Calancala cores from Guajira Region,Colombian Caribbean.

Sample type Laboratory number Depth(cm)

Age* Calibrated Age**

NQ -14C MTC-05329 195 2450+80 2515 yr BPNQ -14C Utc Nr-11923 365 5482±51 6280 yr BPCal14C Poz-475 155 125±25 1880 ADCal 14C Poz-476 199 220±25 1655 ADCal 210Pb My Core-661 1–4 52 2002 ADCal 210Pb My Core-662 11–15 50 2000 ADCal 210Pb My Core-663 21–23 35 1985 ADCal 210Pb My Core-664 26–30 32 1982 ADCal 210Pb My Core-665 36–38 26 1976 ADCal 210Pb My Core-666 41–45 −21 1929 AD

*BP non-calibrated ages for 14C samples and 210Pb.** Calibrated ages for 14C samples and dates reported by the laboratory for 210Pb.

97L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

Fig. 5. Comparison of pollen percentages diagrams of mangroves from analyzed sediment cores at the Colombian Caribbean at two time scales. A. Mangrove changes since4000 cal yr BP to the present, B. Mangrove changes since 450 cal yr BP.

Fig 6. Comparison between summary pollen percentages of salt marshes, Cyperaceae, and mangroves with sample scores of the first axis, obtained by Detrended CorrespondenceAnalysis (DCA). A. Navío Quebrado core, B. Calancala Core. In both cores a good correspondence between DCA scores curve and salt marsh curves, especially Cyperaceae is observed.Mangrove curves present the opposite to DCA scores.

98 L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

resembles other pollen records from the Caribbean (Urrego et al.,2009b), and in fact processes analogous to the one proposed herehave been inferred from sedimentary records of Venezuela and Guy-ana around 7000 yr BP (Van der Hammen, 1988; Rull et al., 1999).Salt marsh expansion was interrupted during a short interval(5730–5170 cal yr BP) that was characterized by rather high propor-tions of Amaranthaceae and Rhizophora, probably reflecting an incip-ient mangrove under direct tidal influence. High percentages ofAmaranthaceae on coastal environments are a good indicator of di-rect tidal flooding (González and Dupont, 2009b) as they are earlycolonizing species that first occupy bare zones of lower and middlemarsh areas with a high incidence of waves and prolonged inunda-tion regimes (Ranwell, 1972). The vegetation sequence along theoldest part of the core probably represents landward migration ofshorelines caused by a rapid sea-level rise, widely reported for thewestern Caribbean during this time interval (Ramcharan, 2004;Wooller et al., 2004, 2007; Ramcharan and McAndrews, 2006;González and Dupont, 2009a; Monacci et al., 2011). Once a balancebetween salty and fresh water is established and allochthonous sedi-ments are provided, biomass tends to increase and true mangrovecommunities flourish (Krauss et al., 2008; Eslami-Andargoli et al.,2009). Thus, increased freshwater input from the high precipitationvalues that characterized the mid-Holocene Thermal Maximum inthe northern Neotropics (Haug et al., 2001) was probably anothercause of the expansion of estuarine environments.

Lower sedimentation rates in Navío Quebrado during the older partof the record were probably the result of low accumulation due to theearly stage of formation of the depositional environment. However, nostrong conclusions can be drawn from Navio Quebrado sedimentationrates as the chronology is based on just two radiocarbon dates. Current-ly, the Navío Quebrado lagoon is enclosed by a coastal plain surroundedby a higher level of Neogene sediments and receives sporadic alluvialsediment input from the Camarones River (Vernette, 1996). It is likelythat during its incipient development stage, the sand barrier that closesthe lagoon (sensu Raasveldt and Tomic, 1958) at its northern part wasnot completely consolidated (Fig. 1). Nevertheless, the mechanismsthat led to the development of the lagoon seem to have acted synchro-nously at regional scale, as this same process has been reported for othersites in the Caribbean (Vernette, 1996; Ramcharan and McAndrews,2006; Peros et al., 2007; Martínez et al., 2010).

The accumulation of organic remains from 3840 to 1270 cal yr BP inNavio Quebrado (Zone NQ II) suggests a depositional environment iso-lated from direct tidal erosive action, probably through the formationof a sand spit that enclosed the lagoon (Fig. 1). These local conditionscombined with regional sea-level stabilization (Rull et al., 1999), andhigh precipitation levels in northern South America (Haug et al.,2001), facilitated the establishment of mangrove communities.Mangrove growth requires marine still stands (Ellison and Stoddart,1991), availability of sheltered tidal sites (Carvalho do Amaral et al.,2006), fresh water supply from either precipitation or river flows forsalinity control (Eslami-Andargoli et al., 2009), alluvial sedimentinput (Krauss et al., 2008) and the autochthonous biotic process of or-ganic accumulation (Mckee et al., 2007). Apparently these conditionscoincided through the period from 3840 to 1270 cal yr BP (Zone NQII), allowing successful establishment of mangrove communities inthe Navío Quebrado area. A similar process of mid-Holocene mangrovecolonization and establishment has been reported for several sites inthe Caribbean. Nevertheless, time frames vary from between ~8000and 4800 cal yr BP (Monacci et al., 2009) to between ~4000and 1700 cal yr BP (Cohen and Wiedemann, 1973; Plaziat andAugustinus, 2004; Peros et al., 2007; Ellison, 2008). This asynchronyin mangrove colonization through the Caribbean reflects the impor-tance of local physiographic settings, as opposed to sea-level regionallycontrolling mangrove dynamics by itself.

From 2850 to 2450 cal yr BP (Zone NQIIB), reduction of man-groves dominated by Rhizophora and expansion of the stands

dominated by Avicennia, the highest salt tolerance of mangrove spe-cies from the study area, indicates hypersalinization of the substrate.Given that during this time interval high sea levels prevailed in theColombian Caribbean (Robertson and Martínez, 1999), the mostlikely explanation for this process is low precipitation and minimumalluvial water input. In fact, dry conditions during this period havebeen reported for the Tesca lagoon and the Cariaco Basin (Haug etal., 2001; Martínez et al., 2010). Additionally, during the dry seasonthe Caribbean Low Level Jet intensifies the regional trade winds,which in these drought times could have promoted the formation ofa permanent sand spit through the deposition of eolic sediments,isolating the lagoon from the salinity-ameliorating tidal action(Ruiz-Ochoa and Bernal-Franco, 2009).

Between 2450 and 1270 cal yr BP (Zone NQIIC), the recovery ofRhizophora, Laguncularia, Batis and Poaceae revealed an expansionof the mangrove forest, probably favored by increased precipitationand marine still stands. During mid-to-late Holocene, local sea levelshigher than present have been reported for the Colombian Caribbean(Robertson and Martínez, 1999), which would conflict with the pres-ence of in situ well-developed mangrove forests (>70% of mangrovepollen) in Navío Quebrado and in Ciénaga Grande de Santa Marta inthe apparent absence of local subsidence (Cohen and Wiedemann,1973). However, our reconstruction shows that the expansion of suit-able areas for mangrove colonization was associated with deltaprogradation and/or regression, rather than with the balancebetween relative sea-level change and local subsidence, as it hasbeen recorded in other places (Ellison, 2008).

Around 2000 yr BP a mangrove decline was evident in NavíoQuebrado, even though at this time strong tidal influence in thelagoon was revealed by presence of shells in the sediments. Suchdecline could be explained by a gradual decrease in precipitationthat has been reported for the northern tropics during the lateHolocene, probably resulting from decreased Northern Hemispheresummer insolation (Hodell et al., 1991; Wooller et al., 2007). Thus,in spite of stable sea levels (Toscano and Macintyre, 2003), reducedprecipitation probably increased salinity, causing a mangrove declinefrom 1270 to 150 cal yr BP (Zone NQIII). Total disappearance ofRhizophora and expansion of salt marsh vegetation in Navío Quebradooccurred synchronously (1240–720 cal yr BP) with low mangrovepollen percentages in the lowermost pollen zone of the Calancalacore (Figs. 5 and 6), reflecting a fringing mangrove forest surroundingsalt marshes. These changes were probably associated with a diminu-tion of alluvial sediments input caused by periods of extreme season-ality and drought in the Caribbean during the Medieval Warm Period(MWP) (Haug et al., 2001; Hodell et al., 2001). By contrast, increasingLaguncularia percentages around 1270 cal yr BP were probably theresult of increased human disturbance reported for the area aroundthis time interval (Ardila, 1996). A similar process is taking placetoday at Navío Quebrado lagoon, where the surface area isundergoing a significant forest reduction due to the confluence ofhigh evaporation rates and increased human disturbances.

Synchronous low percentages of mangrove pollen between 450and 150 cal yr BP in Calancala and Navio Quebrado suggest a contrac-tion of mangrove forest area. These changes were likely the result of aprogressive drought that has taken place in the Circum-Caribbean fol-lowing the MWP (Hodell et al., 1991; Carrillo-Bastos et al., 2010). Infact, the Little Ice Age (LIA) in the area was associated with extremedroughts (Haug et al., 2001; Cohen et al., 2005) that probably causedhypersalinization and deterioration of the environmental conditionsfor mangrove communities. Increments of salt marsh elements(Amaranthaceae, Cyperaceae and Poaceae) in the Colombian Caribbe-an during this time interval might indicate the development ofextended saline mudflats suitable for these species, suggesting eitherthe progradation of the shore after stabilization of sea-level(González and Dupont, 2009b) or the response to the slight incre-ments of sea-level that have been reported for the MWP (Vermeer

99L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

and Rahmstorf, 2009). A similar broadening of the area covered bysalt marsh vegetation has been recorded in Ciénaga Grande de SantaMarta in the Colombian Caribbean (Cohen and Wiedemann, 1973).As the magnitude of estimated sea-level change was relatively smallto produce such significant ecological changes, it is most likely thatreplacement of mangroves by salt marshes revealed by the pollenrecord was the result of coastal progradation.

Mangrove expansion recorded in Calancala from 150 cal yr BP topresent (Zone CALIII) was synchronous with the hiatus of the last146 yr in Navío Quebrado. These contrasting responses were proba-bly associated with geomorphological differences that translatedinto differences in fresh water and continental sediments input. Infact, whereas sedimentation rate at Navío Quebrado is comparablewith those of some Caribbean islands with low allochtonous sedi-ment input (e.g. Ellison, 1993), at Calancala sedimentation rateslargely exceeded other continental mangrove records (e.g. Rull etal., 1999; Ellison, 2008). The difference in sedimentation rates likelyresulted from differences in catchment area of each site as the basinof the river that transports sediments into Navío Quebrado is 10times smaller than that of Calancala (Vélez et al., 2002). Additionally,permanent fresh water input in Calancala, and therefore high sedi-mentation rates (~0.98 cm/yr), facilitates the progradation of thesediments and the subsequent expansion of mangrove communities.

A mangrove expansion similar to that of Calancala for the last150 yr has been reported for the delta of Sinú River, southwesternColombian Caribbean coast (Figs. 1 and 5). High sediment inputfrom river tributaries and sea-level rise gave place to the infilling ofthe bay that favored mangrove expansion in the area (Castaño et al.,2010). High alluvial sediment load diminishes water depth and atten-uates wave impact, benefiting mangrove expansion (Plaziat andAugustinus, 2004). Contrastingly, in Navío Quebrado, extreme sea-sonality of alluvial input, coastal erosion, and low sedimentationrates (0.08 cm/yr) caused lost of substrate and mangrove diminution,as reported for other Caribbean sites (Cohen and Wiedemann, 1973;Ellison, 1993, 2008).

In San Andrés Island, an environmental setting with lowallochtonous sediment input, mangroves have expanded during thelast 300 yr, probably due to wetter conditions at the end of the LIAand onwards (González et al., 2010). In these environments, expan-sion depends on below ground peat accretion (McKee et al., 2007;Ellison, 2008) and freshwater input. Accretion rates in San Andrésmangroves were also high during the last 150 yr (0.42 cm/yr), receiv-ing the necessary freshwater input for their development from sea-sonal water currents (Urrego et al., 2009a). Rates of sea-level riserecorded in the Colombian Caribbean during the last 50 yr (0.36–0.56 cm/yr, Torres et al., 2006) are within the limits of those reportedas maximum tolerable by mangrove communities (0.2–1 cm/yr)(Ellison, 2008). Therefore, high fluvial sediment input and peat accre-tion rates have played a major role in the survival of mangroves byoffsetting the high sea-level rise rates in the region. Conversely,Navío Quebrado lagoon provides an example of the implications oflow sedimentation rates and coastal erosion in the disappearance ofmangrove communities (Correa et al., 2005; Rangel and Anfuso,2009).

5.2. Mangroves: beyond sea level

Local and regional factors play important roles in the establishmentand survival of mangrove communities (Krauss et al., 2008), posing in-trinsic difficulties in the reconstruction of a unique mangrove-basedHolocene sea-level curve. Our data reiterates the importance of geo-morphology, precipitation rates, and fluvial sediment input as determi-nant factors that modulate vegetation response to sea level changes.Thus, spatial interpolation among local reconstructions has to takeinto account all these factors. Since pollen percentages of both saltmarsh and mangrove vegetation show a strong relationship with

local conditions (Roe and Van de Plassche, 2005; Urrego et al., 2009b,2010), analysis of pollen curves of individual mangrove species inconjunction with those of salt marsh vegetation provide the best indi-cation of sea-level changes and environmental processes that occurin coastal plains. For example, salt marsh successional sequences(e.g. Amaranthaceae, Poaceae, Cyperaceae) might be a better proxyfor accelerated sea-level rise than mangroves. DCA scores statisticallysupport these assertions as most of the data variability (DCA Axis 1)seems to mark the divide between mangrove vegetation and saltmarshes (Fig. 6). According to salt marsh pollen spectra we can assertthat between 6000 and 4000 yr BP there was a rapid sea-level rise inthe Colombian Caribbean, coinciding with reports from other Caribbe-an sites and South America (Rull et al., 1999; Ramcharan, 2004; Milneet al., 2005; Carvalho do Amaral et al., 2006). Only with sea-level stabi-lization in the Caribbean around 3000–2500 yr BP (Rull et al., 1999;Behling et al., 2001), mangrove communities established and expanded(Ellison, 2008). However, high precipitation levels that took placewithin this time interval (Haug et al., 2001) were also determinant asthey allowed increased sediment supply, favoring the expansion ofsuitable areas for mangrove colonization. Contrastingly, droughtsrecorded episodically during the MWP and the main part of the LIAprobably played a major role in mangrove area reductions and compo-sitional changes due to hypersalinization.

Our data suggest important changes in sea-level during the last200 yr. Nevertheless, the magnitude of sea-level rise is difficult toassert since tidal flooding levels are confounded with pluvial, geo-morphologic, and fluvial dynamics, making difficult a clear-cut recon-struction of sea-level history. The current marine transgressive eventstarted around 150 yr BP (Kemp et al., 2011) and has been uniformneither in time nor in space. It is in general causing erosion and shore-line retreat (Correa et al., 2005; Rangel and Anfuso, 2009), and asa consequence mangroves die back (Guerrero et al., 2005). However,in small to medium size river deltas, infilling bays, and peat accumu-lating islands mangroves have expanded (Correa et al., 2005; Castañoet al., 2010; González et al., 2010). Another important feature causedby these forest dynamics is the transition frommixed mangrove com-munities to monospecific stands of either Avicennia or Rhizophora,depending on local conditions. Current increasing sea levels posean immense threat to the survival of mangrove communities, espe-cially in the Colombian Caribbean given high current erosion rates(Rangel and Anfuso, 2009) and low alluvial sediment inputs. Unlessthere are coastal plains with high fresh water input, high sedimentsupply (e.g. river deltas and estuaries), and enough space to migrate,coastal erosion will cause extensive mangrove die-back, especiallyin locations that face direct wave action. In fact, it has been suggestedthat mangroves have gone through frequent crises of local ex-tinction and recolonization during Quaternary sea-level fluctua-tions (Woodroffe and Grindrod, 1991). In spite of coastal erosionand the loss of mangrove cover during high sea-level rise phases,recolonization has occurred on continental coastlines with healthy sed-iment budgets.

6. Conclusions

Pollen records from Navío Quebrado and Calancala allowed the re-construction of the history of mangroves and salt marshes in LaGuajira, Colombian Caribbean. The synchronicity of sea level and pre-cipitation changes in La Guajira with other Caribbean records sincethe middle Holocene suggests regional changes, although it is clearthat mangrove dynamics have been strongly influenced by local geo-morphology. Around 6000 cal yr BP, high sea-levels promoted theformation of several lagoons in the Caribbean, Navío Quebradoamong them, which were dominated by salt marsh vegetation.Since then, marine still stands and high precipitation rates havepromoted mangrove colonization in the circum-Caribbean. The dryperiod between 2850 and 2450 cal yr BP reported for the Caribbean

100 L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

probably caused substratum hypersalination and vegetation turnoverfrom Rhizophora- to Avicennia-dominated mangrove stands. Around2500 yr BP, high precipitation levels led to the expansion ofRhizophoramangroves through increases of sediment supply that off-set sea-level rise. Mangroves decline since 2000 yr BP was probablyassociated with a progressive freshwater input diminution inducedby the MWP and the LIA. The response of mangroves to the most re-cent transgressive phase, which started at around 150 yr BP, seems todiffer profoundly depending on the accretion–erosion balance, whichdepends on local conditions. Under sea-level rise scenarios,mangroves are likely to survive if they have high fresh water andsediment supply (e.g. river deltas and estuaries) as was the case inCalancala location. Contrastingly, locations that face direct wave ac-tion and lack permanent fresh water input (such as Navío Quebrado)are much more vulnerable to sea-level rise and result in the extensiveloss of mangroves forests. Overall, our results show that even thoughmangrove communities are assumed to linearly respond to sea level,many factors have to be taken into account when interpretingmangrove-derived pollen sequences.

Acknowledgments

The authors are grateful to Direccion de Investigaciones de laUniversidad Nacional de Colombia, Sede Medellín (DIME) for finan-cial support, Laboratorio de Ecología César Pérez Figueroa for labora-tory logistics, Universidad de la Guajira and Jaime Polanía for theirsupport during field work. Henry Hooghiemstra and Juan D. Lara pro-vided valuable suggestions that strengthened the manuscript.

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.palaeo.2012.11.023.

References

Andrade, C., Barton, E.D., 2005. The Guajira upwelling system. Continental ShelfResearch 25, 1003–1022.

Ardila, G., 1996. Los tiempos de las conchas. Bogotá, ed.Universidad Nacional deColombia, 249 pp.

Behling, H., Cohen, M.C.L., Lara, R.J., 2001. Studies on Holocene mangrove ecosystem dy-namics of the Braganca Península in north-eastern Pará, Brazil. Palaeogeography,Palaeoclimatology, Palaeoecology 167, 225–242.

Carrillo-Bastos, A., Islebe, G.A., Torrescano-Valle, N., González, N.E., 2010. Holocenevegetation and climate history of central Quintana Roo, Yucatán Peninsula, Mexico.Review of Palaeobotany and Palynology 160, 189–196.

Carvalho do Amaral, P.G., Ledru, M.P., Ricardi, F., Fonseca, P.C., 2006. Late Holocenedevelopment of mangrove ecosystem in southern Brazil (Itanhaém, State of SaoPaulo). Palaeogeography, Palaeoclimatology, Palaeoecology 241, 608–620.

Castaño, A.R., Urrego, L.E., Bernal, G.R., 2010. Dinámica del manglar en el complejolagunar de Cispatá (Caribe colombiano) en los últimos 900 años. Revista deBiología Tropical 58, 1347–1366.

Cohen, A.D., Wiedemann, H.U., 1973. Distribution and depositional history of somePre-lagoonal Holocene sediments in the Ciénaga Grande de Santa Marta, Colombia.Boletín de Investigaciones Marinas y Costeras 7, 139–154.

Cohen, M.C.L., Behling, H., Lara, R.J., 2005. Amazonian mangrove dynamics during thelast millennium: the relative sea-level and the Little Ice Age. Review ofPalaeobotany and Palynology 136, 93–108.

Correa, I.D., Alcántara-Carrió, J., González, D.A., 2005. Historical and recent shore ero-sion along the Colombian Caribbean coast. Journal of Coastal Research 49, 52–57((Proceedings of the 2nd Meeting in Marine Sciences), Special Issue).

Ellison, J.C., 1993. Mangrove retreat with rising sea-level, Bermuda. Estuarine, Coastaland Shelf Science 37, 75–87.

Ellison, J.C., 2008. Long term retrospection on mangrove development using sedimentcores and pollen analysis: a review. Aquatic Botany 89, 92–104.

Ellison, J.C., Stoddart, D.R., 1991. Mangrove ecosystem collapse during predicted sea-level rise: Holocene analogues and implications. Journal of Coastal Research 7,151–165.

Erdtman, G., 1960. The acetholysis method. A revised description. Svensk BotaniskTidskrift 54, 561–564.

Eslami-Andargoli, L., Dale, P., Sipe, N., Chaseling, J., 2009. Mangrove expansion andrainfall patterns in Moreton Bay, Southeast Queensland, Australia. Estuarine,Coastal and Shelf Science 85, 292–298.

Faegri, K., Iversen, J., 1989. Textbook of Pollen Analysis. Willey, Chichester. (293 pp.).

Giri, C., Ochieng, E., Tieszen, L.L., Zhu, Z., Singh, A., Loveland, T., Masek, T., Duke, N.,2011. Status and distribution of mangrove forests of the world using earth obser-vation satellite data. Global Ecology and Biogeography 20, 154–159.

González, C., Dupont, L., 2009a. Tropical salt marsh succession as sea-level indicatorduring Heinrich events. Quaternary Science Reviews 28, 939–946.

González, C., Dupont, L., 2009b. Coastal vegetation evidence for sea level changes asso-ciated with Heinrich events. PAGES News 17, 70–72.

González, C., Urrego, L.E., Martínez, J.I., Polanía, J., Yokoyama, Y., 2010. Mangrovedynamics in the southwestern Caribbean since the ‘Little Ice Age’: a history ofhuman and natural disturbances. The Holocene 20, 849–861.

Guerrero, L.V., Argüelles, D., Gomez, I.S., Cordoba, S., Jimenez, N.L., Manrique, J.D., 2005.Santuario de Fauna y Flora Los Flamencos - Plan de Manejo 2005–2009. ParquesNacionales Naturales de Colombia- Ministerio de Ambiente y desarrolloTerritorial- Embajada Real de los Paises Bajos.

Haug, G.H., Hughen, K.A., Sigman, D.M., Peterson, L.C., Röhl, U., 2001. Southward migra-tion of the Intertropical Convergence Zone through the Holocene. Science 293,1304–1308.

HIMAT, 1994. Calendario Meteorológico, 1994. Instituto Colombiano de Hidrología,Meteorología y Adecuación de Tierras, Santafé de Bogotá.

Hodell, D., Curtis, J., Jones, G., Higuera-Gundy, A., Brenner, M., Binford, M.W., Dorsey, K.,1991. Reconstruction of Caribbean climate change over the past 10,500 years.Nature 352, 790–792.

Hodell, D.A., Brenner, M., Curtis, J.H., Guilderson, T., 2001. Solar forcing of droughtfrequency in the Maya Lowlands. Science 292, 1367–1370.

Hooghiemstra, H., Lézine, A.M., Leroy, S.A. Dupont L., Marret, F., 2006. Late Quaternarypalynology in marine sediments: a synthesis of the understanding of pollendistribution patterns in the NW African setting. Quaternary International 148,29–44.

IGAC (Instituto Geográfico Agustín Codazzi), 2009. Estudio general de suelos yzonificación de tierras del departamento de La Guajira. Imprenta Nacional deColombia, Bogotá . (490 pp.).

Kemp, A.C., Horton, B.P., Donnelly, J.P., Mann, M.E., Vermeer, M., Rahmstorf, S., 2011.Climate related sea-level variations over the past two millennia. Proceedings ofthe National Academy of Sciences of the United States of America 108,11017–11022.

Krauss, K.W., Lovelock, C.E., McKee, K.L., Lopez-Hoffman, L., Ewe, S.M., Sousa, W.P.,2008. Environmental drivers in mangrove establishment and early development:a review. Aquatic Botany 89, 105–127.

Lema, L.F., Polanía, J., Urrego, L.E., 2003. Dispersión y establecimiento de las especies demangle del río Ranchería en el período de máxima fructificación. Revista de la Ac-ademia Colombiana de Ciencias Exactas. Físicas y Naturales 27, 93–103.

Lugo, A.E., Snedaker, S.C., 1974. The ecology of mangroves. Annual Review of Ecologyand Systematics 5, 39–64.

Martínez, J.I., Yokoyama, Y., Gomez, A., Delgado, A., Matsuzaki, H., Rendon, E., 2010.Late Holocene marine terraces of Cartagena region, southern Caribbean: the prod-uct of neotectonism or a former high stand in sea-level? Journal of South AmericanEarth Sciences 29, 214–224.

McKee, K.L., Cahoon, D.R., Feller, I.C., 2007. Caribbean mangroves adjust to risingsea-level through biotic controls on change in soil elevation. Global Ecology andBiogeography 16, 545–556.

Medina, E., Cuevas, E., Popp, M., Lugo, A.E., 1990. Soil salinity, sun exposure, and growthof Acrostichum aureum, the mangrove fern. Botanical Gazette 151, 41–49.

Milne, G.A., Long, A.J., Bassett, S.E., 2005. Modelling Holocene relative sea-level obser-vations from the Caribbean and South America. Quaternary Science Reviews 24,1183–1202.

Monacci, N.M., Meier-Grünhagen, U., Finney, B.P., Behling, H., Wooller, M.J., 2009. Man-grove ecosystem changes during the Holocene at Spanish Lookout Cay, Belize.Palaeogeography, Palaeoclimatology, Palaeoecology 280, 37–46.

Monacci, N., Behling, H., Finney, B., Meirs-Groenhagen, U., Wooller, M.J., 2011. Paleo-ecology of mangroves along the Sibun River, Belize. Quaternary Research 76,220–228.

Muller, J., 1959. Palynology of recent Orinoco delta and shelf sediments. Micropalaeontology5, 1–32.

Peros, M.C., Reinhardt, E.G., Davis, A.M., 2007. A 6000-year record of ecological andhydrological changes from Laguna de la Leche, north coastal Cuba. QuaternaryResearch 67, 69–82.

Plaziat, J.C., Augustinus, P.G., 2004. Evolution of progradation/erosion along the FrenchGuiana mangrove coast: a comparison of mapped shorelines since the 18th centu-ry with Holocene data. Marine Geology 208, 127–143.

Raasveldt, H.C., Tomic, A., 1958. Lagunas colombianas. Contribución a la Geomorfologíade la costa del mar Caribe con algunas observaciones sobre las Bocas de Ceniza.Revista de la Academia Colombiana de Ciencias Exactas. Fisicas y Naturales 10,175–198.

Ramcharan, E.K., 2004. Mid-to-late Holocene sea-level influence on coastal wetlanddevelopment in Trinidad. Quaternary International 120, 145–151.

Ramcharan, E.K., McAndrews, J.H., 2006. Holocene development of coastal wetland atMaracas bay, Trinidad, West Indies. Journal of Coastal Research 22, 581–586.

Rangel, N., Anfuso, G., 2009. Assessment of coastal vulnerability in La Guajira PeninsulaColombian Caribbean Sea. Journal of Coastal Research 56, 792–796 (SI).

Ranwell, D.S., 1972. Ecology of Salt Marshes and Sand Dunes. Chapman and Hall(258 pp.).

Robertson, K., Martínez, N.J., 1999. Cambios de niveles del mar Durante el Holoceno enel Litoral Caribe Colombiano. Cuadernos de Geografía 8, 168–199.

Roe, H.M., Van de Plassche, O., 2005. Modern pollen distribution in a Conneticutsaltmarsh: implications for studies of sea-level. Quaternary Science Reviews 24,2030–2049.

101L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102

Author's personal copy

Ruiz-Ochoa, M., Bernal-Franco, G., 2009. Variabilidad estacional e interanual del vientoen los datos del reanálisis NCEP/NCAR en la cuenca Colombia, mar Caribe. Avancesen Recursos Hidráulicos 20, 7–20.

Rull, V., Vegas-Vilarrubia, T., Espinosa de Pernia, N., 1999. Palynological record of anEarly-Mid Holocene mangrove in Eastern Venezuela. Implications for Sea-levelrise and disturbance history. Journal of Coastal Research 15, 496–504.

Sánchez-Páez, H., Álvarez-León, R., Pinto-Nolla, F., Sánchez-Alferez, A.S., Pinto-Rengifo,J.C., García-Hansen, I., Acosta-Peñaloza, M.T., 1997. Diagnóstico y zonificaciónpreliminar de los manglares del caribe de Colombia. Ministerio de MedioAmbiente/OIMT, Santa Fé de Bogotá . (511 pp.).

Stevens, P.F., 2001. Angiosperm Phylogeny Website. Version 12, July 2012 http://www.mobot.org/MOBOT/research/APweb/.

Ter Braak, C.J., Smilauer, P., 2002. CANOCO reference manual and CanoDraw forWindows user's guide: software for canonical community ordination (ver. 4,5).Microcomputer Power, Ithaca . 500 pp.

Tomlinson, P.B., 1995. The botany of mangroves. Cambridge Tropical Biology Series.Cambridge University Press . (436 pp.).

Torres, R.R., Gómez, J.C., Afanador, F., 2006. Variación del nivel medio del mar en elCaribe Colombiano. Boletín Científico CIOH 24, 64–72.

Toscano, M.A., Macintyre, I.G., 2003. Corrected western Atlantic sea-level curve for thelast 11,000 years based on calibrated 14C dates from Acropora palmata frameworkand intertidal mangrove peat. Coral Reefs 22, 257–270.

Urrego, L.E., Polanía, J., Buitrago, M.F., Cuartas, L.F., Lema, A., 2009a. Distribution ofmangroves along environmental gradients on San Andres Island (ColombianCaribbean). Bulletin of Marine Science 85, 27–43.

Urrego, L.E., Bernal, G., Polanía, J., 2009b. Comparison of pollen distribution patterns insurface sediments of a Colombian Caribbean mangrove with geomorphology andvegetation. Review of Palaeobotany and Palynology 156, 358–375.

Urrego, L.E., González, C., Urán, G., Polanía, J., 2010. Modern pollen rain in mangrovesfrom San Andres Island, Colombian Caribbean. Review of Palaeobotany andPalynology 162, 168–182.

Van der Hammen, T., 1988. South America. In: Huntley, B., Webb III, T. (Eds.), Vegeta-tion history. Kluwer, Dordrecht, pp. 307–337.

Vásquez, J., 2000. Estructura de los bosques de manglar del Valle de Los Cangrejos y ElRiíto delta del río Ranchería, Caribe Colombiano. Tesis Ingeniería Forestal.Universidad Nacional de Colombia, Sede Medellín. Medellín.

Vélez, J.I., Mesa, O.J., Poveda, G., Borja, A., Ramírez, J.M., Hernández, O., Urzola, J.A.,Arias, P.A., 2002. HIDROSIG Java: una herramienta para la visualización, análisis yestimación de información hidrológica. XV Seminario Nacional de Hidráulica eHidrología. Sociedad Colombiana de Ingenieros, Universidad Nacional deColombia, Sede Medellín. Medellin.

Vermeer, M., Rahmstorf, S., 2009. Global seal level linked to global temperature. PNAS106, 21527–21532.

Vernette, G., 1996. Ambientes costeros y nivel medio del mar. Sedimentación Marina,Aspectos básicos y aplicados con referencia especial a los litorales colombianos.X Seminario Nacional de Ciencia y Tecnología. Colciencias, Comisión colombianade Oeceanografía, Uinversidad EAFIT.

Woodroffe, C.D., Grindrod, J., 1991. Mangrove biogeography: the role of Quaterna-ry environmental and sea-level change. Journal of Biogeography 18 (5),479–492.

Wooller, M.J., Behling, H., Smallwood, B.J., Fogel, M., 2004. Mangrove ecosystem dy-namics and elemental cycling at Twin Cays, Belize, during the Holocene. Journalof Quaternary Science 19, 703–711.

Wooller, M.J., Morgan, R., Fowell, S.J., Behling, H., Fogel, M., 2007. A multi-proxy peatrecord of Holocene mangrove paleoecology from Twin Cays, Belize. The Holocene17 (8), 1129–1139.

102 L.E. Urrego et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 370 (2013) 92–102