Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

journal homepage: www.elsevier.com/locate/palaeo

Middle Pleistocene palaeolimnology of a dammed tropical river: The ZarzalFormation, Cauca Valley, Colombia

Daniel Jaramilloa,⁎, Diego Felipe Vallejoa,b, María Isabel Vélezc, Sergio Restrepo-Morenod,e,Andres Pardo-Trujilloa,b, Raúl Trejos-Tamayoa,b, Hugo Murciaa,b, Min Kyoungwone,Angel Barbosa-Espitiaa,e

a Instituto de Investigaciones en Estratigrafía (IIES), Universidad de Caldas, Manizales, Colombiab Departamento de Ciencias Geológicas, Universidad de Caldas, Manizales, Colombiac Department of Geology, University of Regina, Regina, CW 234.6, Canadad Departamento de Geociencias y Medio Ambiente, Universidad Nacional de Colombia, Medellín, Colombiae Department of Geological Sciences, University of Florida, Gainesville, FL 32611, USA

A R T I C L E I N F O

Keywords:South AmericaMicropalaeontologyDiatomsNorthern AndesQuaternary

A B S T R A C T

The Plio/Pleistocene Zarzal Formation (ZF) offers one of the few examples of alternation between fully lacustrineto fluvial conditions in a major, tropical, intramontane river in the Northern Andes, i.e., the Cauca River. The ZFis mainly composed of diatomites, mudstones, and tuffaceous sandstones deposited in an intramontane de-pression located between the Western and Central cordilleras of Colombia and associated with the Cauca-Romeral Fault System. Stratigraphic, sedimentological, micropalaeontological (diatoms), and geochronologicalanalyses (independent (U-Th)/He dating of apatite and zircon) have been performed on two sections of the ZF.These analyses reveal a fluctuation between volcanic, fluvial and lacustrine deposits during the MiddlePleistocene. In addition, the presence of soft-sediment deformation structures reflects the synergetic effect oftectonism and volcanism on the sedimentation patterns in the Cauca River Valley. Our data show that betweenabout 0.48–0.52 Ma (Ionian) depositional facies varied from a fluvial-dominated environment to a palaeo-lake,suggesting the impoundment of the river. Once the lake was formed, full lacustrine conditions were frequentlyinterrupted by river pulses bringing terrigenous and previously deposited volcanic material into the lake.Diatom-based palaeolimnological reconstructions and comparisons with the sedimentological records indicatethat river episodes played a decisive role on nutrient supply and sediment input, controlling important ecologicaldynamics in the newly formed lake. Although we cannot unambiguously pinpoint to the influence of volcanicactivity on nutrient cycling within the lake, our data exhibit a strong relationship between volcanic deposits andchanges in flora. The latter are possibly linked to the increase in phosphorous availability. The combined effectsof tectonism and volcanism had major environmental implications in the Quaternary evolution of the CaucaRiver causing unprecedented damming of a one of the major torrential tropical rivers in northern South America.Therefore, are of utmost importance when considering risk assessment and management for riverine commu-nities and infrastructure such as dam buildings.

1. Introduction

The Zarzal Formation (ZF) is a formal name given by Van derHammen (1958) to a set of sedimentary layers composed of diatomites,mudstones, and tuffaceous sandstones outcropping near the town ofZarzal in the Cauca Valley, Central Andes of Colombia (Fig. 1A, B) (seealso De Porta, 1974). Recent studies have suggested that these rockswere formed under a dynamic depositional environment including la-custrine, floodplain, and fluvial conditions, implying damming episodes

of one of the largest rivers in the Northern Andes (Neuwerth et al.,2006; Neuwerth, 2012; Suter et al., 2005, 2008a,b). Although the totalthickness of the ZF is still unknown, Zuñiga and Padilla (1993) andNivia (2001), based on stratigraphic relationships and geophysicalstudies, have estimated that it can reach up to 30 m. This unit is part ofthe Cauca Valley sedimentary fill, which has been influenced by vol-canism from the Central Cordillera and tectonism related to the con-tinued collision of the Panamá-Chocó Bock (Fig. 1B) (Suter et al.,2008b). The Cauca depression records important fluvial activity in an

http://dx.doi.org/10.1016/j.palaeo.2017.08.034Received 9 March 2017; Received in revised form 22 August 2017; Accepted 23 August 2017

⁎ Corresponding author.E-mail address: danieljlo@hotmail.com (D. Jaramillo).

Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

Available online 26 August 20170031-0182/ © 2017 Elsevier B.V. All rights reserved.

MARK

Inter-Andean setting probably since the Eocene (Marín-Cerón et al.,2015; Sierra et al., 2004). At present, the Cauca River runs through adeep intramontane depression located between the Western and Centralcordilleras. With a total length of ~1400 km, a basin area of~63,300 km2, and mean annual discharge in excess of 2000 m3/s, theCauca River is a major tropical fluvial system, it is considered thesecond largest river in Colombia, (Restrepo and Syvitski, 2006;Ocampo-Duque et al., 2013). It ranks among the most densely popu-lated fluvial basins in South America hosting 25% of the Colombian

population settled in this valley (Pérez-Valbuena et al., 2015). Thisstudy aims at unraveling the Middle Pleistocene palaeoenvironmentalhistory of the Cauca River, focusing on the lacustrine-dominated de-posits of two outcropping sections close to Zarzal town (Fig. 1C). Re-sults presented here constitute the oldest record of continental diatomsreported from Colombian geological units.

Fig. 1. Location and structural information of the studied area (A). Spatial distribution of the Zarzal Formation and its relationship with surrounding Quaternary units (B). Geological mapmodified from Neuwerth et al. (2006).

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

195

2. Geological framework

The southern portion of the Cauca Valley is bounded by a series offaults with a predominant N-S trend that correspond to the Cauca-Patíaand Romeral Fault systems (Fig. 1B). The basin itself has been inter-preted as the result of a graben structure (McCourt, 1984; Droux andDelaloye, 1996) or as a transtensional pull-apart basin (Kellogg et al.,1983) (Fig. 1). The studied area is situated where kinematics of theRomeral Fault System change from dextral to sinistral strike-slip due tothe active deformation linked to the collision of the Panamá-ChocóBlock indenter (Suter et al., 2008b). The ZF is part of the oldest Qua-ternary sedimentary infill of the Cauca Basin (Fig. 1), and lies on anangular unconformity over the Miocene fluvial-volcaniclastic depositsof La Paila Formation (McCourt, 1984; Schwinn, 1969; Van derHammen, 1958). Earlier sedimentological studies indicate that the ZFwas deposited under fluvial and lacustrine conditions as well as underintense seismic activity (Cardona and Ortiz, 1994; Neuwerth et al.,2006; Suter et al., 2005, 2008a). The origin of this unit has been con-nected to the downstream damming of the Cauca River. Two mechan-isms have been proposed for this damming: 1) progressive westwardmigration of volcaniclastic mass flows derived from the activity of theSan Diego-Cerro Machín volcanic province in the Central Cordillera (cf.Martínez et al., 2013); these flows, which form the Cartago and Quindiofans, possibly blocked the Cauca River (Fig. 1C) (Cardona and Ortiz,1994; Suter et al., 2005; Neuwerth et al., 2006); and 2) tectonism in-duced by uplifting and indentation of older rocks mainly from theWestern Cordillera that served as a physical barrier in the central partof the Cauca basin (James, 1986; Suter et al., 2008b). Damming periodsin the north of the ZF have also been reported during the Holocene(Page and Mattsson, 1981; García et al., 2011), most likely as a result oftectonism and climate (Schumm et al., 2000; Martínez et al., 2013). Theage of the formation of this lacustrine facies is still a matter of con-troversy, ranging between ~2.8 Ma, based on laser fusion 40Ar/39Ardating of biotites from ash layers (Neuwerth, 2012) and< 0.8 Ma,according to the presence of Alnus (Van der Hammen andHooghiemstra, 1997; Neuwerth, 2012).

3. Materials and methods

Two field sections (S1 and S2) located ~550 m apart have beendescribed, mapped and sampled. Outcrops were cleaned and descrip-tions made on fresh exposure. Sections S1 and S2 are ~12.5 m and 7 mthick, respectively (Fig. 2A). Samples for diatoms were taken at leastevery 25 cm. The most representative samples for volcaniclastic de-posits analysis were collected and the thinner layers described in thefield. Lithofacies have been grouped according to Miall (1996). A det-rital sample from a volcaniclastic layer was collected at the base of S1 toprovide geochronological constrain on the depositional age through (U-Th)/He dating in apatite and zircon.

3.1. Diatom extraction

A total of 84 samples were prepared for diatom extraction by takingaliquots of 300 μl of a solution composed of 0.3 g of dried sedimentsdissolved in 30 ml of distilled water with a few drops of hetamex-aphosphate of sodium (10%). The aliquot was poured on a coverslip,placed on a petri-dish, and left to settle at room temperature for 12 h.This procedure was also followed in order to create a photographicrecord under a Scanning Electron Microscope (FEI QUANTA 250). Aminimum of 400 valves were counted under a Nikon Eclipse 55i-1000× microscope. Taxonomical identifications are based on Tayloret al. (2007), Renan et al. (2012), Bicudo and Menezes (2006), andonline databases Spaulding et al. (2010) and http://craticula.ncl.ac.uk/EADiatomKey/html/index.html. Diatom preservation was defined byvisual estimation and given a number ranging from 0 = very poor,1 = poor, 2 = moderate to 3 = good. Diatoms per gram were

calculated following Flores and Sierro (1997). Relative abundancediagrams have been generated using Tilia 1.7.16® software. Clusteranalysis of the diatom information has been carried out with CONISS®(Grimm, 1987). Graphic correlation of the two studied sections is basedon lithological and biostratigraphical data according to Shaw's method(Shaw, 1964).

3.2. Volcanic deposits

Volcaniclastic deposits were analyzed for each sections followingMurcia et al. (2013). This classification depends on the origin of thedeposits. Pyroclastic fall deposits (ash fall) include sediments formed bydirect accumulation from a pyroclastic fall. Secondary volcaniclasticdeposits include reworked volcanic sediments removed by exogenousprocesses such as gravity, water, air, and/or ice (e.g., lahars).

3.3. Age control

Young volcanic samples often represent a challenge for radiometricdating (Davidson et al., 2004). Typical geochronometers such as U/Pbor C-14 can determine ages either older than ~5 Ma (U/Pb) or youngerthan ~50 ka (C-14). Furthermore, not all sedimentary and volcani-clastic deposits contain datable organic matter suitable for C-14 datingof Quaternary materials (Noller et al., 2000). Recently, the (U-Th)/Hemethod has been successfully applied to dating young volcanics, i.e.,over the Holocene-Pleistocene interval (Min et al., 2006; Farley et al.,2002; Danisik et al., 2012; Cox et al., 2012). From field evidence andprevious age constraints, the expected eruption age for the samplestudied here is< 5 Ma. Therefore, the (U-Th)/He method for datingapatite and zircon crystals has been applied in this study to provide ageochronological constrain for the ZF.

Our age control has been derived from a sample collected 2.2 mabove the base of the S1 section in the ZF, an ash fall deposit (biozoneA; Fig. 2). We followed the standard sample preparation and analyticalprocedures for (U-Th)/He dating (Farley et al., 2002; Reiners, 2005;Restrepo-Moreno et al., 2009). Apatite and zircon separates were ob-tained through conventional gravimetric and magnetic susceptibilityprocedures. Euhedral, unfractured, apatite, and zircons grains with aminimum prism thickness of ~65 μm and low elongation ratios weremanually selected under a high resolution stereomicroscope. Zircon andapatite grains with similar typologies (i.e., overall form, dominantcrystalline faces, and color) were selected to avoid potential mixture ofdifferently sourced-materials. Individual grains were scanned for mi-neral and/or fluid inclusions under a petrographic stereoscope atmagnification 160×. Selected grains for aliquot mounting were digi-tally measured to generate morphometric parameters for FT corrections(Farley et al., 1996; Farley, 2000). After grain selection and measure-ments, single and multigrain aliquots of apatite and zircon were pre-pared. Zircon aliquots were wrapped in Nb tubes preparing a total of sixaliquots, two multigrain and four single grains. For apatite, because ofthe expected young age of the sample and the potentially low U-Th-Smconcentrations, we packed one multigrain aliquot consisting of 10 in-dividual crystals with similar dimensions/shapes in order to improveyields of 4He and U-Th-Sm signals. All apatite crystals were placed intoa single Pt tube. Wrapped sample packets were transferred to a stainlesssteel planchette for full degassing via heating individual packets in anautomated, all-metal analytical system under high vacuum conditions.Helium isotopic ratios were measured using a Pfeiffer-Blazers Prisma®quadrupolar mass spectrometer. Degassed packets were retrieved fromthe planchette to dissolve apatite and zircon grains and measure parentisotope concentrations (235U, 230Th and 149Sm) from a spiked solutionusing a Thermo Finnigan Element2® ICP-MS. Ages were calculatedusing the HelioPlot software® (Vermeesch, 2010). Alpha-recoil FT cor-rections were performed following Farley et al. (1996). Errors are re-ported at 1σ level including both analytical and FT corrections as part oferror calculations. Durango apatite (Farley, 2000) and Fish Canyon Tuff

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

196

zircon (Reiners, 2005) were used as monitoring standards for apatiteand zircon, respectively. These standards were measured every nineunknowns. Procedural hot and cold blanks were routinely measuredduring the sequence of analysis. The standard age equation used for (U-Th)/He dating is based on an assumption that the intermediatedaughter isotopes in the U-Th decay chain have reached secular equi-librium. However, for a very young volcanic sample (<~1 Ma), thisassumption is invalid, requiring a more complex age calculation. Forsuch samples, the correct (U-Th)/He ages are commonly older thanthose calculated from the standard equation by a few tens of percentagedepending on age, D230, etc. (Cox et al., 2012; Danisik et al., 2012). Forthis correction, we have used the equation of Farley et al. (2002) withan assumption that our zircon samples have D230 value (definition) of0.2 which is the average of terrestrial zircon samples. More detailedanalytical procedures are available in Electronic Supplementary Mate-rial.

4. Results

4.1. Lithofacies and diatoms

Five diatom biozones were identified in section S1 (A–E) (Fig. 3),whereas section S2 contains only the three upper biozones (C–E)(Figs. 2 and 4). Lithological descriptions are provided for each of thediatom biozones.

4.1.1. Biozone A (0–2 m, section S1)Thick tabular beds marked by thin of interlayering of siltstone la-

minae and very fine sandstones with parallel, lenticular, and climbingripple laminations. Convolute bedding and water escape structures,including dish-and-pillar, can be locally observed (Subzone A1; Fig. 3).In the upper segment of this interval, there is a ~ 70 cm thick massivebed of beige mudstone with abundant root traces and small wood logs(Subzone A2; Fig. 3). Diatoms consist of benthic and aerophil speciesincluding (average relative abundance in parenthesis): Achnanthidiumminutissimum (23,4%) (Fig. 5K), Diadesmis confervacea (21,6%)(Fig. 5G), Nitzschia amphibia (9%) and Staurosira construens (7,5%) the

0

1

2

3

4

5

6

7

8

9

10

11

12

Biozone D

Biozone C

Biozone B

Biozone A

Biozone E

S1

A

S2

2

3

4

5

6

7

1

0

Thi

ckne

ss (

m)

Thi

ckne

ss (

m)

0

1

2

3

4

5

6

7

8

9

10

11

12

LAD A. ambigua

FAD D. stelligera

FAD S. construens

LAD A. granulata var. angustissima

FAD S. cf. minutulus LAD D.stelligera

LAD A. muzzanensisLAD S. cf. minutulus

LAD M. undulata

S1

S2B

2 3 4 5 6 710

Thi

ckne

ss (

m)

Thickness (m)

Pyroclastic fall depositVolcaniclastic deposits

Cross beddingSedimentary structures

Lithology

Wood fragments

Muddy sandstone Mudstone volcaniclastic sandstones Diatomite

Root traces Sismites Parallel lamination

Load casts

Secondary volcaniclastic deposit dated layer

Mudcracks Graded bedding

Fig. 2. Correlation of section 1 (S1) and section 2 (S2) for the Zarzal Formation. A) Direct stratigraphic correlation between S1 and S2. B) Biostratigraphical and lithological correlationusing Shaw's method. Note the 45°-slope tendency, revealing a very good correlation between S1 and S2. LAD = Last appearance datum, FAD = First appearance datum, Redstar = Geochronological age (~0.5 Ma). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

197

latter peaks in this interval. Planktonic Aulacoseira ambigua (Fig. 5A) isalso common (10,3%). Low abundance of diatoms precluded the esti-mation of valve concentration.

4.1.2. Biozone B (2–5 m, section S1)This biozone records a sharp change in lithology and diatom com-

position (Fig. 3). The lithology consists of thick tabular beds of struc-tureless white diatomites interbedded with thin beds and laminae ofgray, fine to coarse grained sandstones, rich, primarily, in hornblende,pyroxene, and plagioclase crystals, and pumice. It is interpreted as anash fall and secondary volcaniclastic deposits, which can displaynormal grading. Mud cracks and sand dykes intruding the diatomitelayers are also identified at the base of this biozone (Fig. 3). Diatomsassemblages are composed of planktonic species A. ambigua and A.granulata (Fig. 5B) with averages of 50,6% and 41,9% respectively, andsome appearances of Fragilaria crotonensis as well as planktonic Dis-costella stelligera (Fig. 5D) at the top (Fig. 3). Valve concentration is of8 × 108 valves/g (Fig. 3).

4.1.3. Biozone C (5–7.6 m, section S1; 0–2.8 m, section S2)This is the lowest biozone registered in both sections (Figs. 2 and 4).

The lithology is similar to that of the previous interval, but volcani-clastic deposits become less frequent (Fig. 3). In terms of diatom as-semblages A. ambigua disappears from the record while other species ofAulacoseira such as A. pusilla (18%), A. granulata v. angustissima(18,2%), and A. muzzanensis (7,4%) become more abundant (Fig. 3).Planktonic diatoms F. crotonensis is present with a maximum of 7%, andD. stelligera shows peaks of high abundances (max 40%) (Fig. 3). Thehighest diatom concentration occurs in this biozone with values up to3 × 109 valves/g.

4.1.4. Biozone D (7.6–10.4 m, section S1; 2.8–5.3 m, section S2)The lithology of this interval is characterized by an increase in thin

diatomite and mudstone interbeds. Seven thin beds of gray sandstoneseither massive or with normal grading can be observed. According tothe composition and structure, we have classified these layers as sec-ondary volcaniclastic (particularly at the base) and pyroclastic fall de-posits (Fig. 3). Particles of Fe oxide, which are concordant with thestratification, are common in this zone. Diatomites are mixed withdetrital sediments and several samples lack diatoms (Fig. 3). Dish-and-pillar and soft-sediment deformation structures are common and arehelpful features for correlation between volcaniclastic deposits in bothsections (Fig. 2). The diatom signal includes the first and last occur-rence of Stephanodiscus cf. minutulus (Fig. 5C) (average 11,4%, max30%), and abundant S. construens (max 19%) and S. pinnata (max16,5%) (Figs. 3 and 4). Aulacoseira spp. decreases (57%) and D. stelligeradisappears completely from the record. Diatom concentration is re-duced substantially to 3 × 108 valves/g (Fig. 3).

4.1.5. Biozone E (10.4–12.5 m, section S1; 5.3–7 m, section S2)This interval is made largely of laminated mudstones which are

interbedded with some diatomite laminae. Two thin beds of volcani-clastic sandstones can be observed at the top (Fig. 3). A. granulata(73,8%), S. construens (14,1%), and S. pinnata (8,2%) dominate thisbiozone. Diatom production shows a decline (~1.3 × 108 valves/g),and evident fragmentation of Aulacoseira valves is observed (Figs. 3 and4).

4.2. Radiometric dating

The new (U-Th)/He data obtained from apatite and zircon samples

0

1

2

3

4

5

6

7

8

9

10

11

12

Sedi

men

tary

str

uctu

res

Vol

cani

clas

tic d

epos

its

S1

PlanktonicLife-Habits

Legend

Pyroclastic fall depositVolcaniclastic deposits

Cross beddingSedimentary structures

Lithology

Wood fragments

Muddy sandstone Mudstone volcaniclastic sandstones Diatomite

Root traces Sismites Parallel lamination

Load casts

Secondary volcaniclastic deposit dated layer

Facultative planktonic Benthic Aerophiles

Barren samples

Mudcracks Graded bedding

Thi

ckne

ss (

m)

CONISS

4 6 8 10 12 14 16 18 20 22

Total sum of squares

BIOZONE

Life-H

abit

Diatom

s

A1

A2

B

C

D

E

203E+90 40 60 80 100 20 40 60 80 100 20 40 20 20 20 40

Pres

erva

tion

N=

Val

ve/g

ram

s

Aul

acos

eira

gra

nul

ata

A. a

mb

igua

A.

gran

ulat

a va

r. a

ngu

stis

sim

a

A.

muz

zan

ensi

s

A.

pusi

lla

Step

hano

disc

us c

f. m

inut

ulus

20 2020

Fra

gila

ria

cro

ton

ensi

s

Stau

rosi

ra c

onst

ruen

s

Stau

rosi

rell

a pi

nnat

a

20 40

Dis

cote

lla s

telli

gera

20 20 20 20 40 60

Mel

osir

a un

dula

ta

Cyc

lote

lla m

eneg

hin

ian

aSy

ned

ra u

lna

Coc

cone

is p

lace

ntu

laN

itzs

chia

am

phib

ia

Ach

nant

hid

ium

min

utis

sim

um

A.

exig

uum

Nav

icul

a s

ppP

lano

thid

ium

fre

quen

tiss

imum

Fra

gila

ria

capu

cina

Sella

phor

a p

upul

aP

innu

lari

a sp

pF

rust

ulia

spp

Nit

zsch

ia a

cido

clin

ata

Nit

zsch

ia s

aca

lari

sE

uno

tia

mai

or

Go

mp

hone

ma

turr

isG

om

pho

nem

a p

arv

ulum

En

cyon

ema

spp

Rho

palo

dia

spp

Dia

desm

is c

onfe

rvac

ea

Luti

cola

geo

ppe

rtia

naLu

tico

la m

uti

caN

avic

ula

ga

llica

v. l

aevi

ssim

aO

rth

osei

ra d

endr

ote

res

Dia

desm

is c

onte

nta

Eu

noti

a p

ecti

nal

is v

. un

dula

ta

20 40 60 80 100

(%)

Fig. 3. Biostratigraphical distribution of diatoms and the biozones according to Coniss and defined biozones. Lithology, sedimentary structures, and interpretation of volcaniclasticdeposits in section S1. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

198

are available in Tables 1 and 2, respectively. One multigrain apatitealiquot (10 apatite grains; Toba Ap) yielded corrected (U-Th)/He age of2.1 ± 0.3 Ma with Th/U ratio of 2.5. Two multigrain zircon aliquotsgenerated corrected dates of 1.9 ± 0.2 (Toba Zr_Multi 1) and0.60 ± 0.08 Ma (Toba Zr_Multi 2). Four out of five single grain zirconaliquots yielded corrected ages ranging from 0.20 ± 0.38 to0.67 ± 0.47, all of which are internally consistent within 2-sigmauncertainty. These ages are virtually identical to one multigrain zirconaliquot (Toba Zr Multi 2; 0.60 ± 0.08 Ma). The five single grain zirconages and one multigrain zircon age (Toba Zr Multi 2) generate a

weighted mean of 0.53 ± 0.06 Ma with MSWD of 0.5.

5. Discussion

5.1. Geochronology and stratigraphical correlations

Except for two multigrain aliquots of apatite (Toba Ap,2.1 ± 0.4 Ma at 2σ), and zircon (Toba Zr_Multi 1, 1.8 ± 0.2 Ma at2σ), all the remaining aliquots yielded consistent (U-Th)/He ages in therange 0.20–0.60 Ma with a weighted mean of 0.53 ± 0.12 Ma at 2σ

2

3

4

5

6

7

1

0

Thi

ckne

ss (

m)

0 14E+8

Pres

erva

tion

N=

Val

ve/g

ram

s

Au

laco

seir

a a

mb

igua

A.

gran

ulat

a

A.

gran

ulat

a v.

an

gust

issi

ma

A.

muz

zan

ensi

s

A.

pusi

lla

Step

hano

cicl

us c

f. m

unut

ulus

Fra

gila

ria

cro

ton

ensi

s

Mel

osir

a un

dula

ta

Stau

rosi

ra c

onst

ruen

sSt

auro

sire

lla

pinn

ata

Cyc

lote

lla m

eneg

hin

ian

a

Dis

cote

lla s

telli

gera

Nit

zsch

ia a

mph

ibia

20 40 60 80 100

Vol

cani

clas

tic d

epos

its

Sedi

men

tary

str

uctu

res

S2

PlanktonicLife-Habits

Legend

Pyroclastic fall depositVolcaniclastic deposits

Sedimentary structures

Lithology Muddy sandstone Mudstone volcaniclastic sandstones Diatomite

Sismites Parallel lamination

Secondary volcaniclastic deposit dated layer

Facultative planktonic Benthic Aerophiles Barren samples

Graded bedding

Life-H

abit D

iatom

s

20 40 60 80 100 20 20 20 40 20 40 601 2 3 4

Total sum of squares

CONISSBIOZONE

D

C

E

(%)

Fig. 4. Biostratigraphical distribution diagrams of diatoms with Coniss analysis and their biozones. Lithology, sedimentary structures, and interpretation of volcaniclastic deposits insection S2.

Fig. 5. Common diatom species. A. A. ambigua, B. A. granulata, C. Stephanodiscus cf. minutulus, D. D. stelligera, E.Melosira undulata, F. Orthoseira dendroteres, G. D. confervacea, H. Cocconeisplacentula, I. Staurosira construens, J. Staurosirella pinnata, K. A. minutissimum.

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

199

and. We suggest that the latter represents the most likely timing oferuption for this volcanic unit and consequently that our sample datesback to the Middle Pleistocene (Ionian stage) (Gibbard et al., 2010).This new and independent chronological estimate for the ZF is inagreement with an age of< 0.8 Ma based on palynological observa-tions, i.e., the presence of Alnus (Hooghiemstra and Sarmiento, 1991;Andriessen et al., 1993; Van der Hammen and Hooghiemstra, 1997;Suter et al., 2008a; Neuwerth, 2012), but differs from previously re-ported 40Ar/39Ar Gelasian ages for this same unit (Neuwerth, 2012).

The two multigrain aliquots (Toba Ap_Muilti and Toba Zr_Multi 1)with significantly older ages than the rest of our aliquots probablycontain “exotic” detrital grains from nearby layers. This demonstratesthe importance of single grain analysis particularly for sedimentary andvolcaniclastic samples. Despite efforts to select un-reworked apatite andzircon grains with unvarying typologies, the grains within these ali-quots may not be exclusively derived from the grains originating fromthe corresponding eruption, but from a mixture of grains from previousvolcanic and tectonic episodes. It remains unclear why two of theanalyzed multigrain aliquots yielded an identical age within their 2σuncertainties (~2 Ma). One possible explanation is that the 0.5 Maeruption episode was predated by an additional 2 Ma volcanic eruptionand that the level sampled in this study could be made by a mixture ofash fall-out of the Ionian volcanic event plus the contribution of someGelasian and Calabrian (Early Pleistocene) grains from previous events.

In the temporal context provided by our helium ages, sections S1and S2 are correlated using lithological features, cluster analyses,diatom biozones, and a 45° slope in the Shaw's graph (Fig. 2B). BiozonesC, D, and E are present in the two sections (Figs. 3 and 4) displaying anhomogenous extensive lacustrine environment that spread out for atleast 550 m. Alternations between volcanic-dominated and fluvial-la-custrine deposits, as well as the presence of soft-sediment deformationstructures, reflect the synergetic effects between Cauca River dynamics,tectonism, and volcanism during the Middle Pleistocene (~0.5 Maago).

5.2. Palaeolimnological and palaeoecological reconstruction

Biozone A corresponds to an overbank deposit (Fig. 3) under theeffect of river pulses as indicated by the alternation of fine-grainedsandstones and mudstones with cross-bedding and planar lamination insubzone A1. In addition, epipelic A. minutissimum and epiphytic C.placentula suggest presence of macrophytes, most likely reflectingshallow conditions. The peaks of A. ambigua occurring in sedimentsfrom fluvial environments of moderate energy (e.g. cross-beddedsandstones) are indicating that this species came via the Cauca River(Fig. 3). A. granulata and A. ambigua are considered as riverine speciescommonly found in floodplain lakes (Hay et al., 2000; Gell et al., 2002;Reid and Ogden, 2009). High abundance of aerophil D. confervaceaoccurring in massive mudstones with abundant root traces reveals theformation of ephemeral ponds isolated from the main flow in subzoneA2 (Fig. 3).

In biozone B, diatom assemblage was dominated by planktonicspecies A. ambigua and A. granulata. This assemblage, along with theappearance of pure diatomite layers, and the absence of benthic

diatoms, indicate a shift to permanent lacustrine conditions (Fig. 3).These Aulacoseira species are common in floodplain shallow lakes witha dynamic mixing regime, high concentrations of silica and high tur-bidity (Hotzel and Croome, 1996; Liu et al., 2012). The high abundanceof A. ambigua in the lower part of the zone and is decrease in the upperpart of biozone B (Fig. 2) indicates a gradual decrease in river influence.A subaerial period is revealed by mud cracks at 3.5 m with no obviouschanges in the lithological record (Fig. 3), suggesting that shallowconditions and bottom lake exposure occurred. After this subaerial in-terval, high water levels returned as it is indicated by the abundance ofA. granulata, D. stelligera, F. crotonensis and A. pusilla (upper part ofbiozone B and biozone C) (Fig. 3). D. stelligera and F. crotonensis, whichare good indicators of abundant nitrogen (N) and limited phosphorus(P) (Saros et al., 2011; Saros and Anderson, 2015), suggest high waterlevels rich in these elements. Moreover, F. crotonensis is known to in-crease when silica content (Si) is moderate to high (Saros et al., 2011).

In biozone D, an increase in volcanic activity is inferred from theseveral layers of pyroclastic material product of direct fall and volca-niclastic flows (Fig. 2A). The sudden increase of S. minutulus, Staurosiraconstruens, and S.pinnata, right on top of the volcaniclastic deposits(Figs. 3 and 4), evidences new limnological conditions, most likely as aresult of volcanism affecting the physical and chemical characteristicsof the water. Volcanism can alter the chemistry of the water and/or canform barriers for water circulation (Barker et al., 2003; Solovieva et al.,2015). Both species of Staurosira are known for being aggressive colo-nizers (Urrutia et al., 2010), thus suggesting a change from previousecological conditions. The increase of S. minutulus is normally asso-ciated with waters rich in P (Wang et al., 2012), most likely indicates anenrichment in P. In higher latitude lakes that suffer thermal stratifica-tion, S. minutulus is more abundant with the spring increases in P, butits appearances in tropical waters, where seasonal variations are notimportant, is a matter of debate. According to Vilaclara et al. (2010), itsappearance is correlated with strong perturbation in the ecosystem as aresult of volcanism, while Solovieva et al. (2015) have explained theincrease of dissolved silica from tephras as one of the possible causes forS. minutulus appearance. In Zarzal Formation, the abundance of Aula-coseira spp. throughout the record, suggest that silica was alwaysabundant and available, since this species required high concentrationsof it (Kilham et al., 1996). So far, in freshwater environments nomodern studies have been conducted to understand the relationshipbetween P and volcanism. In marine environments it has been observedthat dissolve P increases during frequent influxes of volcanic materials(Frogner et al., 2001). In the case of the ZF, it is possible that volca-niclastic flows could have brought more P into the system and/or al-tered circulation, enhancing remobilization of P, but more studies arenecessary to find out the effects of volcanism in the trophic state oflakes. Following this quiet period, the Cauca River started flooding thelake once again (biozone E) as suggested by the presence of abundantterrigenous components, the physical fragmentation of valves, and theoccurrence of benthic and facultative planktonic diatoms. Continuousfluvial discharges most likely diluted nutrients which negatively af-fected the previous flora (S. minutulus) and favored the establishment ofa new one (Fig. 3). In summary, a higher production of diatom valveswas recorded in the lacustrine facies when the interaction with the river

Table 1Summary of apatite (U-Th)/He data for sample TOBA.

Sample 4He [fmol] 1σ [fmol] U [fmol] 1σ [fmol] Th [fmol] 1σ [fmol] Sm [fmol] 1σ [fmol] Th/U(wt/wt)

Mass [g] No. grains FT (U-Th)/He age[Ma]a

(U-Th)/He age[Ma]b

1σ

Durango 130 4.8 546 9.2 12,200 127.7 14,204 169.6 21.8 1 1.0 30.0 1.1TOBA Ap_Multi 0.6 0.1 243 3.2 615 8.2 7322 132.8 2.5 9.5E−6 10 0.6 2.1 2.1 0.3Durango 108 4.0 509 6.4 9658 124.6 11,068 157.4 18.5 1 1.0 30.6 1.2

a U-disequilibrium uncorrected.b U-disequilibrium corrected.

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

200

was minor, and the proportion of Si and most probably the ratio of N:Pincreased.

5.3. Origin of the lacustrine conditions in the Zarzal Fm

Our results reveal that lake conditions were reached about 0.5 Maago (Figs. 1 and 3). Reported damming episodes of the Cauca River areknown (Suter et al., 2008a,b), and thus seem to be the more plausibleexplanation for the lacustrine environments recorded in sections S1 andS2. Cause (s) for this impoundment are still poorly understood. Al-though our studied sections are located away from the presumed zoneof damming (Fig. 1C), the 30 cm ash fall layer at transition betweenbiozones A and B demonstrates active volcanism at the time of im-poundment (Fig. 3). The ZF is mainly dominated by fine-grainedlithologies with continuous planar-tabular stratification interlayeredwith sporadic coarse-grained sandstones and tuffs derived from thevolcanic activity from the Central Cordillera (Suter et al., 2005, 2008a).Alternatively, these characteristics might indicate that great reliefcontrast generated by tectonic-driven uplift and subsidence controlledmost of the sedimentation through fluvial tributaries generating la-custrine conditions similar to those of ria lakes (Schumm et al., 2000).In fact, structural analyses suggest that tectonism has been importantcontrolling factor on the sedimentation in the Cauca and Quindio-Ri-saralda basins during the Quaternary (Suter et al., 2008b; López et al.,2009). Based on all of these characteristics, we consider that thedamming of the Cauca River most likely involves tectonic and volcanicprocesses creating accommodation space for volcanic and lacustrinesediments to accumulate with sporadic influence of the river. In orderto test this hypothesis, provenance and structural geology studies arerequired, particularly focusing on downstream segments of the CaucaRiver, where potential barrier outcrops might be present.

6. Conclusions

The ZF records a long-term Middle Pleistocene damming of one ofthe most important rivers of the Northern Andes. Damming of thismagnitude is unprecedented for a torrential tropical river. Our studiesbased on diatoms, sedimentary facies, and U-Th/He thermochronologydocument that tectonism and volcanism had major implications in thepre-Quaternary and Quaternary evolution of the Cauca River. Fluvialpulses and volcanic activity had important roles on the ecological dy-namics of the lake by altering nutrient availability, water circulation orlight penetration. Although we cannot unambiguously pinpoint to theinfluence of volcanic activity on nutrient cycling in the lake, our resultsexhibit a strong relationship between volcanic deposits and changes inflora, potentially linked to the increase in phosphorous availability.

This study illustrates that the combined effects of tectonism andvolcanism play an important role in the dynamics of intra-mountainrivers. Consequently, these aspects have to be considered in risk as-sessment exercises for the management of local riverine activities, andcivil infrastructures on the Cauca River.

Acknowledgments

We would like to thank Diego Usma, Susana Velazquez, Julio Avila,Alejandra Gutierrez, and Susana Osorio for their help in the field. JuanPablo Betancurt for his help with graphic work, Universidad EAFIT andUniversidad Nacional de Colombia (Medellin) for facilitating lab spaceand technicians (e.g., mineral-separation master Wilton Echavarria)that were tremendously helpful in sample preparation for heliumdating. This research did not receive any specific grant from fundingagencies in the public, commercial, or not-for-profit sectors.

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.Table2

Summaryof

zircon

(U-Th)/H

eresultsforsampleTO

BA.

Sample

4He[fmol]

1σ[fmol]

U[fmol]

1σ[fmol]

Th[fmol]

1σ[fmol]

Th/U

(wt/wt)

Mass[g]

No.

grains

F T(U

-Th)/H

eag

e[M

a]a

(U-Th)/H

eag

e[M

a]b

1σ

FCT

237

4.9

7680

109.3

3924

56.0

0.51

5.6E

−6

10.8

27.9

–0.7

FCT

1500

30.5

52,907

755

30,466

434.5

0.58

3.6E

−5

10.9

22.8

–0.6

TOBA

ZrMulti1

9.1

0.8

5532

7926

7038

.50.48

4.6E

−6

40.7

1.80

1.90

0.2

TOBA

ZrMulti2

5.1

0.8

10,065

144

5157

73.5

0.51

6.5E

−6

50.7

0.54

0.60

0.08

TOBA

Zrgrain1

2.3

0.8

4893

7123

6833

.70.48

4.1E

−6

10.8

0.43

0.49

0.16

TOBA

Zrgrain3

0.3

0.8

2086

30.2

1182

17.1

0.57

1.5E

−6

10.7

0.14

0.20

0.38

TOBA

Zrgrain6

2.3

0.8

6034

8629

0641

.30.48

3.7E

−6

10.8

0.36

0.42

0.13

TOBA

Zrgrain7

1.0

0.8

1641

2489

012

.80.54

2.3E

−06

10.7

0.61

0.67

0.47

TOBA

Zrgrain8

1.1

0.8

2566

37.2

1359

19.5

0.53

1.5E

−6

10.7

0.46

0.52

0.32

FCT

874

17.8

26,134

371.5

13,546

192.8

0.52

1.8E

−5

10.8

27.7

–0.7

aU-diseq

uilib

rium

unco

rrected.

bU-diseq

uilib

rium

corrected.

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

201

doi.org/10.1016/j.palaeo.2017.08.034.

References

Andriessen, P.A.M., Helmens, K.F., Hooghiemstra, H., Riezebos, P.A., Van der Hammen,T., 1993. Absolute chronology of the Pliocene-Quaternary sediment sequence of theBogota area, Colombia. Quat. Sci. Rev. 12 (7), 483–501.

Barker, P., Williamson, D., Gasse, F., Gibert, E., 2003. Climatic and volcanic forcing re-vealed in 50,000-year diatom record from Lake Massoko, Tanzania. Quat. Res. 60,368–376.

Bicudo, C.E., Menezes, M., 2006. Gêneros de Algas de Águas Continentais do Brasil (chavepara identificação e descrições). RiMa.

Cardona JF and Ortiz M (1994) Aspectos estratigráficos de las unidades del intervaloPlioceno Holoceno entre Pereira y Cartago. Propuesta de definición para laFormación Pereira. Universidad de Caldas. Tesis de grado, pp. 155 unpublished.

Cox, S.E., Farley, K.A., Hemming, S.R., 2012. Insights into the age of the Mono LakeExcursion and magmatic crystal residence time from (UTh)/He and 230Th dating ofvolcanic allanite. Earth Planet. Sci. Lett. 319–320, 178–184.

Danisik, M., Shane, P., Schmitt, A.K., Hogg, A., Santos, G.M., Storm, S., Evans, N.J.,Fifield, L.K., Lindsay, J.M., 2012. Re-anchoring the late Pleistocene tephrochronologyof New Zealand based on concordant radiocarbon ages and combined 238U/230Thdisequilibrium and (U–Th)/He zircon ages. Earth Planet. Sci. Lett. 349–350,240–250.

Davidson, J., Hassanzadeh, J., Berzins, R., Stockli, D., Bashukooh, B., Turrin, B.,Pandamouz, A., 2004. The geology of Damavand volcano, Alborz Mountains,northern Iran. GSA Bull. 116, 16–29.

De Porta, J., 1974. Lexique stratigraphique international scientifique. 5. Centre Nationalde la recherche scientifiqueed, Paris, pp. 4b.

Droux, A., Delaloye, M., 1996. Petrography and geochemistry of Plio-Quaternary calc-alkaline volcanoes of Southwestern Colombia. J. S. Am. Earth Sci. 9, 27–41.

Farley, K.A., 2000. Helium diffusion from apatite: general behavior as illustrated byDurango fiuorapatite. J. Geophys. Res. 105, 2903–2914.

Farley, K.A., Wolf, R.A., Silver, L.T., 1996. The effects of long-alpha-stopping distances on(U–Th)/He ages. Geochim. Cosmochim. Acta 60, 4223–4229.

Farley, K.A., Kohn, B., Pillans, B., 2002. The effects of secular disequilibrium on (U-Th)/He systematics and dating of Quaternary volcanic zircon and apatite. Earth Planet.Sci. Lett. 201, 117–125.

Flores, J.A., Sierro, F.J., 1997. Revised technique for calculation of calcareous nannofossilaccumulation rates. Micropaleontology 43 (3), 321–324.

Frogner, P., Gislason, S.R., Oskarsson, N., 2001. Fertilizing potential of volcanic ash inocean surface water. Geology 29, 487–490.

García, Y.C., Martinez, J.I., Velez, M.I., Yokoyama, Y., Battarbee, R.W., Suter, F.D., 2011.Palynofacies analysis of the late Holocene San Nicolás terrace of the Cauca paleolakeand paleohydrology of northern South America. Palaeogeogr. Palaeoclimatol.Palaeoecol. 299, 298–308.

Gell, P.A., Sluiter, I.R., Fluin, J., 2002. Seasonal and interannual variations in diatomassemblages in Murray River connected wetlands in north-west Victoria, Australia.Mar. Freshw. Res. 53, 981–992.

Gibbard, P.L., Head, M.J., Walker, M.J.C., 2010. Formal ratification of the QuaternarySystem/Period and the Pleistocene Series/Epoch with a base at 2.58 Ma. Quat. Sci.Rev. 25 (2), 96–102.

Grimm, E., 1987. CONISS: a Fortran 77 program for stratigraphically constrained clusteranalysis by the method of incremental sum of squares. Comput. Geosci. 13, 13–35.

Hay, M.B., Michelutti, N., Smol, J.P., 2000. Ecological patterns of diatom assemblagesfrom Mackenzie Delta lakes, Northwest Territories, Canada. Can. J. Bot. 78, 19–33.

Hooghiemstra, H., Sarmiento, G., 1991. Long continental pollen record from a tropicalintermontane basin; late Pliocene and Pleistocene history from a 540-meter core.Episodes 14 (2), 107–115.

Hotzel, G., Croome, R., 1996. Population dynamics of Aulacoseira granulata (EHR)SIMONSON (Bacillariophyceae, centrales), the dominant alga in the Murray River,Australia. Arch. Hydrobiol. 136 (2), 191–215.

James ME (1986) Estudio sismotectónico en el area del Viejo Caldas. Ingeominas,Medellín, pp. 113 unpublished.

Kellogg, J.N., Godley, V.M., Ropain, C., Bermúdez, A., 1983. Gravity Anomalies andTectonic Evolution of Northwestern South America: Transactions 10th Caribbean.Geological Conference, Cartagena, Colombia. pp. 18–31.

Kilham, S., Theriot, E.C., Fritz, S.C., 1996. Linking planktonic diatoms and climate changein the large lakes of the Yellowstone ecosystem using resource theory. Limnol.Oceanogr. 41, 1052–1062.

Liu, Q., Yang, X., Anderson, N.J., Liu, E., Dong, X., 2012. Diatom ecological response toaltered hydrological forcing of a shallow lake on the Yangtze floodplain, SE China.Ecohydrology 5, 316–325.

López, M., Moreno, M., Audemard, F., 2009. Deformación tectónica reciente en los pie demontes de las Cordillera Central y Occidental, Valle del cauca, Colombia. Bol. Geol.31 (1).

Marín-Cerón, M., Restrepo-Moreno, S.A., Bernet, M., Foster, D.A., Min, K., Pardo-Trujillo,A., Barbosa-Espitia, A., Kamenov, G., 2015. The Amagá formation (Eocene?-Miocene)in the Cauca-Patía depression: a recorder of major morphotectonic and paleogeo-graphic events between the Western and Central cordilleras. In: Proceedings GSAAnnual Meeting, Baltimore, Maryland, USA. Geological Society of America Abstractswith Programs 47. pp. 675.

Martínez, J.I., Mayr, C., Velez, M.I., Yokoyama, Y., Battarbee, R., 2013. The San Nicolássuccession of the Cauca paleolake: a late Holocene laminated ria lake record, from theNeotropics. J. Paleolimnol. 49, 287–300.

McCourt, W.J., 1984. The geology of the Central Cordillera in the departments of Valle

del Cauca, Quindío and NW Tolima. In: Report no. 8 INGEOMINAS-British GeologicalSurvey 8 Cali, (37 pp.).

Miall, A.D., 1996. The geology of fluvial deposits. Sedimentary facies, basin analysis, andpetroleum geology. Spring 134 (3), 409–421.

Min, K., Reiners, P.W., Wolff, J.A., Mundil, R., Winters, L., 2006. (U–Th)/He dating ofvolcanic phenocrysts with high-U–Th inclusions, Jemez Volcanic Field, New Mexico.Chem. Geol. 227, 223–235.

Murcia, H.F., Borrero, C.A., Pardo, N., Alvarado, G.E., Arnosio, M., Scolamacchia, T.,2013. Depósitos volcaniclásticos: Términos y conceptos para una clasificación enespañol. Rev. Geol. Am. Central 48, 15–39.

Neuwerth, R., 2012. Lithostratigraphical and tectono-sedimentary study of the Plio-Pleistocene infill of the Interandean North Cauca Valley Basin (Colombia). (Ph.d.thesis) In: Terre 2 Environment 106, Earth and Environment Science Section.University of Genova (157 pp.).

Neuwerth, R., Suter, F., Guzman, C.A., Gorin, G.E., 2006. Soft-sediment deformation in atectonically active area: the Plio-Pleistocene Zarzal Formation in the Cauca Valley(Western Colombia). Sediment. Geol. 186, 67–88.

Nivia A (2001) Mapa geologico del departamento del Valle del Cauca. Escala 1:250000.Memoria explicativa. Ministerio de minas y energia. Instituto de investigación einformación geocientífica, minero-ambiental y nuclear ingeominas.

Noller, J.S., Sowers, J.M., Colman, S.M., Pierce, K.L., 2000. Quaternary Geochronology:Methods and Applications. American Geophysical Unionpp. 582.

Ocampo-Duque, W., Osorio, C., Piamba, C., Schuhmacher, M., Domingo, J., 2013. Waterquality analysis in rivers with non-parametric probability distributions and fuzzyinference systems: application to the Cauca River, Colombia. Environ. Int. 52, 17–28.

Page, W.D., Mattsson, L., 1981. Landslide lakes near Santa Fé de Antioquia. Rev. CIAF 6,469–478.

Pérez-Valbuena, G.J., Arrieta, A.M., Contreras, J.G., 2015. Río Cauca: la geografíaeconómica de su área de influencia. 225. Banco de la República-Cartagenapp.1692–3715.

Reid, M.A., Ogden, R.W., 2009. Factors affecting diatom distribution in floodplain lakesof the southeast Murray Basin, Australia and implications for palaeolimnologicalstudies. J. Paleolimnol. 41 (3), 453–470.

Reiners, P.W., 2005. Zircon (U-Th)/He thermochronometry. Rev. Mineral. Geochem. 58,151–179.

Renan, K., Santos, D.S., Cristina, A., Leite, C., Anna, S., 2012. Diatoms from shallow lakesin the Pantanal of Nhecolândia, Brazilian wetland. Oecologia Australis 16 (4),756–769.

Restrepo, J.D., Syvitski, J.P.M., 2006. Assessing the effect of natural controls and land usechange on sediment yield in a major Andean river: the Magdalena drainage basin,Colombia. AMBIO J. Hum. Environ. 35 (2), 65–74.

Restrepo-Moreno, S.A., Foster, D.A., Stockli, D.F., Parra, L.N., 2009. Long-term erosionand exhumation of the “Altiplano Antioqueño”, Northern Andes (Colombia) fromapatite (U-Th)/He thermochronology. Earth Planet. Sci. Lett. 278, 1–12.

Saros, J.E., Anderson, N.J., 2015. The ecology of the planktonic diatom Cyclotella and itsimplications for global environmental change studies. Biol. Rev. 90, 522–541.

Saros, J.E., Michel, T.J., Interlandi, S.J., Wolfe, A.P., 2011. Resource requirements ofAsterionella formosa and Fragilaria crotonensis in oligotrophic alpine lakes: implica-tions for recent phytoplankton community reorganizations. Can. J. Fish. Aquat. Sci.62, 1681–1689.

Schumm, S.A., Dumont, J.F., Molbrook, J.M., 2000. Active Tectonics and Alluvial Rivers.Cambridge University Press, Cambridge, pp. 276.

Schwinn, W.L., 1969. Guidebook to the geology of the Cali area, Valle del Cauca,Colombia. In: Tenth Annual Field Conference. Colombian Society of PetroleumGeologists and Geophysicists, pp. 277–297.

Shaw, A.B., 1964. Time in Stratigraphy. McGraw Hill, New York, pp. 1–365.Sierra, G., Silva, J., Correa, L., 2004. Estratigrafía Secuencial de la Formación Amagá.

Boletin de Ciencias de la Tierra 15, 9–22.Solovieva, N., Klimaschewski, A., Self, A.E., Jones, V.J., Andrén, E., Andreev, A.A.,

Hammarlund, D., Lepskaya, E.V., Nazarova, L., 2015. The Holocene environmentalhistory of a small coastal lake on the north-eastern Kamchatka Peninsula. Glob.Planet. Chang. 134, 55–66.

Spaulding, S.A., Lubinski, D.J., Potapova, M., 2010. Diatoms of the United States. http://westerndiatoms.colorado.edu, Accessed date: 16 November 2016.

Suter, F., Neuwerth, R., Guzman, C., Gorin, G., 2005. Modelo de depositación para laFormación Zarzal y sus relaciones estratigráficas con el Quaternario fluvio-volcánicoproveniente de la Cordillera Central. X Congreso colombiano de geología, Bogotá,Memorias. pp. 86.

Suter, F., Neuwerth, R., Gorin, G., Guzmán, C., 2008a. (Plio-) Pleistocene alluvial-lacus-trine basin infill evolution in a strike-slip active zone (Northern Andes, Western-Central Cordilleras, Colombia). Geol. Acta 6, 231–249.

Suter, F., Sartori, M., Neuwerth, R., Gorin, G., 2008b. Structural imprints at the front ofthe Chocó-Panamá indenter: field data from the North Cauca Valley Basin, CentralColombia. Tectonophysics 460 (1), 134–157.

Taylor, J.C., Harding, W.R., Archibald, C.G.M., 2007. An illustrated guide to somecommon diatom species from South Africa. In: WRC Report TT, pp. 282.

Urrutia, R., Araneda, A., Torres, L., Cruces, F., Vivero, C., Torrejón, F., Barra, R., Fagel, B.,2010. Late Holocene environmental changes inferred from diatom, chironomid, andpollen assemblages in an Andean lake in Central Chile, Lake Laja (36°S).Hydrobiologia 648, 207–225.

Van der Hammen, T., 1958. Estratigrafía de Terciario y Maestrichtiano continentales ytectogenesis de los Andes colombianos. Bol. Geol. 6, 67–128.

Van der Hammen, T., Hooghiemstra, H., 1997. Chronostratigraphy and correlation of thePliocene and Quaternary of Colombia. The Plio-Pleistocene boundary. Quat. Int. 40,81–91.

Vermeesch, P., 2010. HelioPlot, and the treatment of overdispersed (U-Th-Sm)/He data.

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

202

Chem. Geol. 271 (3–4), 108–111.Vilaclara, G., Martinez-Mekler, G., Cuna, E., Ugalde, E., 2010. Diatom-inferred pa-

laeoenvironmental changes of a Pliocene lake disturbed by volcanic activity. J.Paleolimnol. 44, 203–215.

Wang, P., Shen, H., Xie, P., 2012. Can hydrodynamics change phosphorus strategies of

diatoms? In: Nutrient Levels and Diatom Blooms in Lotic and Lentic Ecosystems.Microbiology of Aquatic Systems 63. pp. 369–382.

Zuñiga, O., Padilla, L.E., 1993. Exploracion geofísica de bentonitas y diatomitas en elcorregimiento de vallejuelo del municipio de Zarzal (VALLE). Rev. Cienc. 10(0121–1935), 41–49.

D. Jaramillo et al. Palaeogeography, Palaeoclimatology, Palaeoecology 487 (2017) 194–203

203