Bergner AGN & Trauth MH (2004) - in-africa.org

20 Pages·2004·3.41 MB·English

Palaeogeography,Palaeoclimatology,Palaeoecology215(2004)17–36
www.elsevier.com/locate/palaeo
Comparison of the hydrological and hydrochemical evolution
of Lake Naivasha (Kenya) during three highstands
between 175 and 60 kyr BP
A.G.N. Bergner*, M.H. Trauth
Institutfu¨rGeowissenschaften,Universita¨tPotsdam,POB601553,Potsdam,Germany
Received6August2003;receivedinrevisedform1July2004;accepted21July2004
Abstract
Three diatomite beds exposed in the Ol Njorowa Gorge south of Lake Naivasha, Central Kenya Rift, document three
major lake-level highstands between 175 and 60 kyr BP. Diatom transfer-function estimates of hydrological and
hydrochemical parameters suggest that a deep and large freshwater lake existed during the highstands at ~135 and ~80 kyr
BP. In contrast, a shallower but more expanded freshwater lake existed at ~110 kyr BP. The best analog for the most
extreme highstand at ~135 kyr BP is the highstand during the Early Holocene humid period from 10 to 6 kyr BP. The
environmental conditions as reconstructed from diatom assemblages suggest long-lasting episodes of increased humidity
during the high lake periods. This contrasts to the modern situation with a relatively shallow Lake Naivasha characterized
by rapid water level fluctuations within a few decades. The most likely cause for the variable hydrological conditions
since 175 kyr BP is orbitally driven insolation changes on the equator and increased lateral moisture transport from the
ocean.
D2004Elsevier B.V. All rightsreserved.
Keywords:EastAfrica;Diatoms;Transferfunctions;Lakelevel;Paleoclimate;Pleistocene
1. Introduction beenintenselystudiedinmodernlakes,andalsofrom
coresanddrysurfaceoutcropsofpaleolakesediments
Diatoms are very sensitive indicators of various all over the world (Taub, 1996; Bradley, 1999;
environmental parameters in a lake, such as water Battarbee, 2000; Gasse, 2000; Owen, 2002). Diatom-
depth,chemistry,turbidityandnutrientsupply(Gasse ite as sediment, comprising mainly fossil diatom
etal.,1995).Typicalassemblagesofthesealgae have skeletons, is a good archive of the conditions in a
lake through time (e.g., Smol et al., 2001). Because
changes of these conditions are mainly linked to
climate, diatom assemblages provide indirect insights
* Correspondingauthor.
into the changes of the precipitation–evaporation
E-mailaddress: [email protected]
(A.G.N.Bergner). balance in a catchment. A quantitative estimate of
0031-0182/$-seefrontmatterD2004ElsevierB.V.Allrightsreserved.
doi:10.1016/j.palaeo.2004.07.033
18 A.G.N.Bergner,M.H.Trauth/Palaeogeography,Palaeoclimatology,Palaeoecology215(2004)17–36
the environmental conditions through time is gained initiated lake basin is bounded by rift escarpments to
using diatom transfer functions (Gasse et al., 1995). theeastandwest,Mt.Eburruvolcanotothenorthand
Although reworking of diatom frustules, selective Mt. Longonot volcano and the Olkaria Volcanic
dissolution of some diatom taxa, and diagenetic Complex to the south. The history of the Naivasha
alteration of the biogenic silica after deposition can basin began at about 320 kyr BP, when Olkaria lava
cause significant distortions of the paleoenvironmen- flowsclosedthebasintothesouthbetweentheflanks
tal signal, cross-validation of the environmental of the 400-kyr-old Mt. Longonot and the escarpment
reconstructions based on diatom assemblages with tothewest(Clarkeetal.,1990;Fig.2).Theseeffusive
other proxies emphasizes the validity of the method bodiesareunconformablyoverlain byanupto60-m-
(Gasse et al., 1997). thick fluvio-lacustrine sequence, between 175 and 60
In the Central Kenya Rift, diatoms have been kyr old, suggesting that the basin was covered by a
studied for a long time to reconstruct environmental lake three times larger than modern Lake Naivasha
changesinlakebasinsandhenceclimatechangesona (Trauthetal.,2001,2003;Bergneretal.,2003).After
varietyoftimescales(i.e.,Nilsson,1931;Richardson the regression of this lake in the Late Pleistocene,
and Richardson, 1972; Richardson and Dussinger, ongoing volcanic activity in the Olkaria Complex
1986; Trauth et al., 2001). As identified in Late producedmostofthepresentreliefcoveringtheolder
Pleistocene and Holocene deposits of Lake Naivasha, lakedeposits(Clarkeetal.,1990).Aftertheformation
diatom-based paleohydrological reconstructions of the Ol Njorowa Gorge by headward erosion of the
allowed to manifest periods of high water levels and outlet of the Early Holocene highstand, the Late
low salinity at around 135, 110, 80 and 9 kyr BP Pleistocene sediments became exposed (Washbourn-
(RichardsonandDussinger,1986;Trauthetal.,2003). Kamau, 1977). Today, the sediments are laterally
The precipitation–evaporation ratio during the most continuousoveradistanceofmorethan7kmandcan
extreme highstands at ~9 and ~135 kyr BP has been be sampled in excellent dry sample outcrops of the
determined using a lake-balance model, suggesting a gorges walls.
minimumlong-term increaseinprecipitationof about The modern Lake Naivasha is unique among the
30% (Bergner et al., 2003). However, the short-term other lakes in the Central Kenya Rift because of its
trends and fluctuations as well as the amplitude of relatively low pH (~7.9) and electric conductivity
hydrochemical fluctuations on time scales on decadal (~250 and 500 ASdcm(cid:1)1; Gasse et al., 1995;
to millennial time scales have not been investigated. Verschuren, 1999). The freshness of the lake can be
We therefore studied the diatom assemblages con- attributed to a significant groundwater seepage
tained in three diatomite deposits exposed in the Ol through permeable volcanic subsurface rocks and to
Njorowa Gorge south of the present lake. Using the perennial freshwatersupply from the Malewa and
diatom transfer functions, we compared three high- Gilgil rivers (Ojiambo and Lyons, 1996; Verschuren,
stands at ~135, ~110 and ~80 kyr BP with both the 1999). The two streams drain a ~3200 km2 large
well-established highstand between 10 and 6 kyr BP catchment area, which includes moist mountain
(Richardson and Richardson, 1972; Washbourn- ranges of the eastern (2200 to 2500 m a.s.l.) and
Kamau, 1975; Richardson and Dussinger, 1986) and western escarpments (up to 4400 m elevation).
themodernconditionsinthelake.Theresultsprovide Whereas annual rainfall exceeds 1750 mmdyear(cid:1)1
valuableinsightsintothenaturalvariabilityofthelake in these regions, the fairly plain and wind-stressed
systemanditsresponsetoregionalandglobalclimate Lake Naivasha basin is characterized by a negative
change. moisture budget with rainfall averaging ~650
mmdyear(cid:1)1 and evaporation of ~1900 mmdyear(cid:1)1
(Ja¨tzold, 1981; Kenya Meteorological Department,
2. Setting 2000).
Temporal variations in the hydrological budget of
Locatedat1890mabovesealevel,LakeNaivasha Lake Naivasha are mainly controlled by changes in
is the highest water body in the Central Kenya Rift precipitation, which is in turn linked to the seasonal
(0855VS 36820VE; Figs. 1 and 2). The tectonically migration of equatorial convergence zones, i.e., the
A.G.N.Bergner,M.H.Trauth/Palaeogeography,Palaeoclimatology,Palaeoecology215(2004)17–36 19
Fig.1.Regionalsettingsofthestudyarea;(A)topographyandlocationoflargelakes;migrationofIntertropicalConvergenceZone(ITCZ,
maximumpositionofJulyandJanuaryindicated)andCongoAirBoundary(CAB,dashedcorridor)causebimodalprecipitationpatternwith
rainyseasonsinApril/May(longrains)andOctober/November(shortrains).(1)LakeTurkana,(2)LakesMagadiandNatron,(3)LakeEyasi,
(4)LakeVictoria,(5)LakeAlbert,(6)LakeTanganyika,(7)LakeMalawi.(B)Airflowpatternofthestudyareachangingsignificantlyafter
transitionoftheITCZandCABwithpredominantwestwindstreamsfromtheCongobasinandtradewindsfromIndianOcean(indicatedby
arrows). Shaded areas refer to regions with high precipitation as indicated. Symbols symbolizing sea-surface temperature anomalies in the
IndianOceanrefertohighermoistureavailabilityduringElNin˜oevents.
Intertropical Convergence Zone (ITCZ) and the Melack, 1981; Tarras-Wahlberg et al., 2002). Geo-
Congo Air Boundary (Nicholson, 1996). Although biochemical processes in the nearshore parts of the
heaviest rainfall occurs during April–May and Octo- lake account for much of the ionic removal and
ber–November after the transition of the ITCZ, therefore may provide further explanation for the low
prevailing SE trade winds and westerly winds during alkalinityofthelake(GaudetandMelack,1981).The
summer,aswellasNEtradewindsandnorthwesterly modernhydrochemistryofLakeNaivashaisclassified
air flow during winter cause minor rainfall in the as low-chloride, high-fluoride, sodium-bicarbonate
Naivasha basin (Fig. 1). The short-term variability in water, where the acquisition of solutes by weathering
the intensity of these wind systems is directly related of the surrounding feldspathoidic bedrocks masks the
to sea-surface temperature (SST) variations in the chemicalcompositionoftherain(GaudetandMelack,
Indian,PacificandAtlanticOceans(Camberlin,1995; 1981;Tarras-Wahlbergetal.,2002).Duetotheheavy
Saji et al., 1999). In particular, the El Nin˜o/Southern afternoon winds and the shallow bathymetry, the
Oscillation (ENSO) accounts for a significant part of water body of Lake Naivasha is polymictic and not
hydrological fluctuations on decadal time scales thermally stratified (Verschuren, 1999). The intense
(Camberlin, 1995; Nicholson, 1996; Indeje et al., mixing of the water column is also reflected in the
2000). modern diatom assemblages, collected from both,
Asaconsequenceofthesevariations,thelakelevel sediment and surface-water (i.e., plankton) samples.
is subjected to significant changes, which particularly Herein, the typical modern diatom flora is dominated
affect the shallow areas of the lake (Gaudet and by Aulacoseira ambigua, Aulacoseira granulata var.
20 A.G.N.Bergner,M.H.Trauth/Palaeogeography,Palaeoclimatology,Palaeoecology215(2004)17–36
A.G.N.Bergner,M.H.Trauth/Palaeogeography,Palaeoclimatology,Palaeoecology215(2004)17–36 21
angustissima and Synedra acus (Richardson and to the total sum of counted valves. Furthermore, we
Dussinger, 1986; Gaudet and Melack, 1981; Gasse documented the occurrence of phytoliths and sponge
et al., 1995). spicules as indicators of nearshore deposition. The
ratio of broken diatoms, the amount of clastic debris
as well as the phytolith content were regarded as a
3. Methods proxyforlittoralconditionsofthesamplingsite(e.g.,
Barker et al., 1990; Gasse et al., 1997). Because
According to an earlier composite chronology, the several sections of the profiles show strong lamina-
diatomite beds of the Ol Njorowa Gorge, south of tions, high-resolution sampling and separate analysis
modern Lake Naivasha, represent marginal deposits wereinvestigatedina15-cm-thickcontinuoussliceof
of three high lakes between 175 and 60 kyr BP diatomite to distinguish between the thin dark brown
(Trauthetal.,2001;Bergneretal.,2003;Fig.2Aand and thicker white laminae.
B). Based on 17 single-crystal 40Ar/39Ar-age deter- The diatom identification followed the principles
minations on intercalated tuff layers, the oldest of Hustedt (1949), Gasse (1986) and Krammer and
diatomites, 340 and 120 cm thick, were accumulated Lange-Bertalot (1991a,b, 1997a,b). For paleohydro-
during highstands IX (~135 kyr BP) and VIII (~110 logical and environmental interpretations, the identi-
kyrBP),whereastheyoungestdiatomite,upto50cm fied taxa were cross-checked with the species listed
thick, records highstand V (~80 kyr BP; Trauth and in the modern East African Diatom Database, where
Strecker, 1996; Trauth et al., 2003). We resampled both taxa counts and environmental variables of the
the most prominent profiles of these deposits for a sampling site are included (Gasse et al., 1995).
high-resolution microfossil analyses (cf. locations at Using this database, diatom-inferred conductivity,
Fig. 2C–F). pH, cation and anion ratios were derived by transfer
Depending on the quality of the sediment, we functions. Secondly, principle-component analysis
sampled at 10- to 30-cm intervals for highstand IX (PCA) and hierarchical cluster analysis (CA) were
and VIII, and at 2.5-cm intervals for highstand V. In used to identify groups of diatoms with comparable
order to average potential seasonal variations in the chemical and habitat characteristics. The routine pca
species assemblages, we integrated 2.5-cm-thick of the MatlabR PLS_Toolbox provided by Eigen-
slices of sediment for the investigation of the long- vector Research, Manson, was used to perform PCA
term trends. The dry sample material was washed in on the auto-scaled data, implying all variables were
distilled water, organic matter was oxidized by H O put on an equal basis in the analysis (Swan and
2 2
and carbonates were removed using HCl. Diatom Sandilands, 1995; Wise et al., 2002). This was
slidesweremountedinNaphraxandanalyzedusinga important because the ecological parameters of the
Leica optical microscope at magnification (cid:2)1000. diatoms show large differences in the absolute values
Becauseofthelimitedpreservationofthediatoms,we of mean and variance. Next, we employed the
countedbetween300and1000valvespersample.We MatlabR routine cluster on the auto-scaled data to
also estimated the content of detrital material and the define groups of diatoms with similar environmental
percentage of broken valves for each sample relative preferences.
Fig.2.(A)MapoftheNaivashabasinshowingpresent-daytopography,extensionofmodernLakeNaivasha,mainrivers,locationofsurface
outcropsintheOlNjorowaGorge(squareindicatedby*)andsedimentcorefromlake(triangleindicatedby**;RichardsonandDussinger,
1986).(B)Present-dayreliefandlakeareaobtainedfromdigital-elevationmodeling.Theidealizedlandscapeignoresvegetationcoverage,but
shows the southern rim of the Naivasha basin as seen towards NE with (1) Lake Naivasha, (2) NNE–SSWorientated Ol Njorowa Gorge,
volcaniccomplexesof(3)Olkariaand(4)Longonot,aswellas(5)Akiraplains.LocationofsedimentsequenceintheOlNjorowaGorge(*)
andsedimentcoreprofilefromLakeNaivasha(**).(C)3.40mofdiatomitesurfaceoutcropintheOlNjorowaGorge,showingintercalated
tephra(hammerforscale)and(D)laminatedsectionsofhigherdetritalcontents.Reconstructionofmaximumextensionofpaleo-LakeNaivasha,
asinferredfromtheresultsofthisstudyandreconstructionofthebathymetrybasedongeologicfielddata(Bergneretal.,2003)showing(E)
relief-sketchofthesouthernrimofthepaleobasinderivedfromdigital-elevationmodelingoutliningtheextensionofthepaleo-LakeNaivasha
andspatialrelationofthesedimentprofilesand(F)estimatedtopographyoftheNaivashabasinwith100-m-contourinterval;1900ma.s.l.
correspondstomaximumlakelevelat~135kyrBP.Labelingcorrespondingto(A)and(B).
22 A.G.N.Bergner,M.H.Trauth/Palaeogeography,Palaeoclimatology,Palaeoecology215(2004)17–36
4. Results the relative importance of facultative–planktonic
diatoms as deep-water indicators in littoral habitats,
4.1. Litho-/biofacies correlation we recalculated the planktonic–littoral ratio by
including the facultative–planktonic taxa in the
The diatomite beds of the Ol Njorowa Gorge plankton group (Fig. 3). Done for each highstand
consist by almost 100% of intact or variably separately, the parameters nicely correlate with the
fragmented diatom valves. Phytoliths and sponge changes in the lithologic facies as outlined above.
spicules appear in subordinate numbers, i.e., less Highstand IX is composed offour stratigraphic zones
than 1% of the total number of counted particles. (Figs. 3 and 4): The lowest part of the diatomite
Glass shards and siliciclastic material are more extending from the baseto 105 cm, comprises littoral
frequent, but are generally restricted to the dark diatoms, such as Epithemia adnata, Epithemia sorex,
layers within the diatomite beds. The quality of the Gomphonema intricatum and Cymbella cistula, but
laminations within the diatomites was found to be an the assemblage is clearly dominated by facultative–
appropriate classification scheme to link macroscopic planktonic species, such as Fragilaria pinnata and F.
characteristics to the microscopic character of the construens. Deep-water indicators, such as Aulaco-
sediments. Consequently, the diatomite beds can be seira granulata, Cyclotella ocellata, Cyclotella
classified into three lithologic facies: (a) pure-white glomerata and Nitzschia tropica are rare and only
diatomite with weak lamination, (b) diatomite with increase slightly in the uppermost section. Within the
distinct lamination and (c) grayish diatomite with a less or nonlaminated zone 2 (105 to 180 cm), the
relatively high clastic component, but without clear portion of the facultative–planktonic Fragilaria spp.
lamination. Microscopic inspection of the sediment is reduced, whereas periphytic Cocconeis placentula,
character and preservation of diatom valves support E. sorex, G. intricatum and Mostogloia elliptica
our macroscopic classification scheme: The preser- become more frequent. Zone 3 (180 to 250 cm) is
vation of the diatoms is inversely correlated with the again laminated and characterized by predominant
numbers of phytoliths and sponge spicules as well as planktonic species, mainly of Aulacoseira spp. and
the clastic particles (Fig. 3). Although the laminated Cyclotella spp. In the uppermost, siltic and non-
sections show large internal variation in these laminated part of the diatomite, again more near-
parameters, they contain predominantly well-pre- shore, littoral taxa prevail, predominated by
served diatom valves and minor amounts of phyto- facultative–planktonic Fragilaria spp. and a higher
liths, sponge spicules and clastic debris. Within the fraction of periphytic E. adnata, Cymbella muellerii,
nonlaminated sections, this pattern is reversed; and Gomphonema gracile (Fig. 4A).
significant amounts of broken valves occur more A similar, but much less distinct trend is observed
frequently, while high numbers of clastic debris, in the diatomite profile of highstand VIII (Figs. 3,
phytoliths and sponge spicules are more abundant. 4B): Planktonic species, like Aulacoseira granulata
Assuming that clastic contaminations, phytoliths, or Nitzschia vanoyei, and facultative–planktonic
sponge spicules and broken valves record more diatoms, such as Cyclotella stelligera and Synedra
turbulent, nearshore sedimentary conditions, whereas ulna, are concentrated in the lower, more laminated
intact diatom valves and the absence of phytoliths section of the profile (zone 1). The maximum in the
and sponges indicate calm, deep-water environments planktonic–littoralratiooccursataround30cmabove
(Hecky and Kilham, 1973; Smol et al., 2001), we thediatomitebase, correspondingtoadecreaseinthe
can use our classification of diatomites to estimate occurrence of phytoliths and sponge spicules. In this
water depth and distance to the lake shore (Fig. 3). section,alsothepreservationofthediatomfrustulesis
In the following, we apply this approach to the full good, and only minor amounts of clastic debris are
thickness of the diatomites as well as to sections on observed. In the upper part of the profile, where this
lamination scale. dominance of littoral genera, such as Fragilaria,
A first-order estimate for the hydro-ecological Gomphonema and Epithemia gets striking, the con-
changesinthelakecanbebestillustratedbytheratio tent of clastic debris as well as the number of
of planktonic vs. littoral species. Taking into account phytoliths and sponge spicules increases. Herein, the
A
.G
.N
.
B
e
rg
n
e
r,
M
.H
.
T
r
a
u
th
/
P
a
la
e
o
g
e
o
g
r
a
p
h
y
,
P
a
la
e
o
c
lim
a
to
lo
g
y
,
P
a
la
e
o
e
c
o
lo
g
y
2
1
5
(2
0
0
4
)
1
7
–
3
6
Fig.3.Lithostratigrapicprofilesofdiatomitesof(A)highstandIX(146F2to~120kyrBP),(B)highstandVIII(113F2to108F7kyrBP)and(C)diatomitesanddiatomaceoussilts
ofhighstandV(81F4to73F3kyrBP).Agecontrolfrom40Ar/39Ardatingonsanidinephenochrystsfromintercalatedvolcanictuffbeds(Trauthetal.,2001,2003).Semiquantitative
plotsofsedimentologicalparametersobtainedfromthin-sectionanalysis.Planktonic–littoralratioofdiatomtaxareflectingthefractionofplanktonicvs.periphyticandfacultative–
planktonicdiatoms(solidline),andplanktonic,includingfacultative–planktonicvs.periphyticdiatoms(dashed)withintheidentifieddiatomassemblages.Environmentalinformation
correspondstothedatasetofEastAfricandiatoms(Gasseetal.,1995).
2
3
2
4
A
.G
.N
.
B
e
rg
n
e
r,
M
.H
.
T
r
a
u
th
/
P
a
la
e
o
g
e
o
g
r
a
p
h
y
,
P
a
la
e
o
c
lim
a
to
lo
g
y
,
P
a
la
e
o
e
c
o
lo
g
y
2
1
5
Fig.4.Relativeabundanceofselecteddiatomspeciesinprofilesof(A)highstandIX,(B)highstandVIIIand(C)highstandV.Shadingscorrespondtosedimentfacies(cf.Fig.3). (20
0
4
)
1
7
–
3
6
A.G.N.Bergner,M.H.Trauth/Palaeogeography,Palaeoclimatology,Palaeoecology215(2004)17–36 25
).
d
e
u
n
nti
o
c
(
4
g.
Fi
26 A.G.N.Bergner,M.H.Trauth/Palaeogeography,Palaeoclimatology,Palaeoecology215(2004)17–36
fraction of planktonic diatoms is low and never was obviously not intense enough to completely
exceeds 8%. destroy the layering of the sediments. The presence
AlthoughthediatomaceousbedsofhighstandVdo of laminae therefore suggests limited oxygen and
not show any laminations, even these units show nutrient supply during deeper-water conditions,
similarities between sediment facies and the relative whereas the absence of laminate documents shallow-
abundanceof diatom taxa (Figs. 3and 4B).Through- water nearshore environments. Similar observations
out the profile, periphytic and facultative–planktonic weredescribedincomparabledepositsofEastAfrican
diatomspredominateinthediatomassemblages.Only lakes (e.g., Roberts et al., 1993; Damnati and Taieb,
within a distinctive, narrow zone between 58 and 64 1995; Gasse and Van Campo, 2001).
cmabovethebaseofthediatomite(zone2),anabrupt Althoughtheunderstandingoftheinternaldynam-
increase in planktonic, freshwater diatoms of Aulaco- ics of the lake system would require a more detailed
seiraambigua(upto40%),accompaniedbyabundant investigationofthelaminatedpartofthesection,these
Aulacoseira granulata and Cyclotella stelligera, is preliminaryresultshelptointerprettheobservedlong-
observed.Interestingly,theamountofclasticdebrisis term trends in the diatom flora. As reflected in the
significantly reduced and phytoliths and sponge planktonic–littoral ratio, all highstands are character-
spicules occur more seldom in this section. ized by long-term trends from more littoral, shallow-
In order to fully understand the relationship water conditions to deeper freshwater environments.
between the occurrence of clastic materials and the This trend is overlain by short-term fluctuations, as
biological inventory, a short section of the well- indicated by the highly varying sedimentological and
laminated highstand-IX diatomite (212 to 225 cm paleontological parameters in the laminated sections
aboveitsbase)wasanalyzedin0.25-cmintervalsand oftheprofile.Inthesesections,littoral,i.e.,periphytic
from thin sections. The comparison of the species and facultative–planktonic species get more abundant
assemblages within pure-white and brownish clay- where brownish silt layers alternate with pure-white
rich layers indicates a clear anticorrelation between diatomite. The relative importance of littoral taxa
the amount of clastic material and the number of inversely correlates with the thickness of the white
planktonicdiatoms.Moreover,decreasingnumbersof layers,asitcanbestbestudiedinthelaminated,basal
phytoliths and sponge spicules correlate with higher part of the highstand-VIII diatomite (Fig. 2D).
amounts of well-preserved diatom frustules and a While interpreting the results from this analysis, it
predominance of planktonic genera in the diatom has to be kept in mind, that the fossil flora may
assemblages (Fig. 5). From thin section analysis, it represent an only incomplete picture of the original
can also be concluded that clay-rich, dark layers do assemblage. The limited preservation of atleast some
not document annual or seasonal layers, but most of the samples suggests that this factor could be of
likely reflect repeated events of higher clastic input. some relevance in the Ol Njorowa Gorge sequence.
Such features can best be explained by rhythmic Aboveall,breakingoflargerNavicula,Pinnularia or
depositionofsuspendedmaterial,whichisinitiatedby Synedra frustules is a problem throughout the
enhanced sediment input or turbulences in the littoral sequences and thin-walled frustules are less abundant
lake areas (Wetzel, 2001). Both processes would thanmorphologicallystrongerdiatoms.Thetreatment
account for a reworking of loose sediment, an ofthedrysamplematerialfordiatomslidesmayhave
increased transport of littoral deposits towards deeper also affected fragile skeletons. However, the compar-
parts of the lake and hence enhanced mixing of ison of diatom slides with smear slides and thin
planktonicandlittoraldiatomfragments.Additionally, sections yields no significant difference in the
bioturbation may disturb the original laminated preservation of the diatom frustules. Obviously, the
texture of lake sediments. Benthic oxygenation and damage of the valves has to be explained by
nutrientsupplyarethekeyparametersinfluencingthe sedimentary processes in the paleolakes. Because the
intensity of benthic mixing in aquatic environments location of the Ol Njorowa Gorge section is believed
(Robbins et al., 1977; Battarbee et al., 2001). to represent the littoral part of a lake, reworking of
Although these is some evidence for tracks and diatomfrustulesmaybeofsomeimportance(Gasseet
burrows in some parts of the diatomites, bioturbation al., 1997).