Palaeogeography, Palaeoclimatology, Palaeoecology 303 (2011) 126 – 132
Contents lists available at ScienceDirect
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p a l a e o
a , ⁎
b
b
a b
Department of Ecology and Evolutionary Biology, The University of Arizona, Tucson, AZ 85721, USA
Department of Geosciences, University of Arizona, Tucson, AZ 85721, USA a r t i c l e i n f o
Article history:
Received 27 February 2009
Received in revised form 23 October 2009
Accepted 1 January 2010
Available online 7 January 2010
Keywords:
Lake Malawi
Cichlid
Fish fossils
Carbon isotopes
Cyprinid
Sediment core a b s t r a c t
Numerous biological and chemical paleorecords have been used to infer paleoclimate, lake level fl uctuation and faunal composition from the drill cores obtained from Lake Malawi, Africa. However, fi sh fossils have never been used to examine changes in African Great Lake vertebrate aquatic communities nor as indicators of changing paleolimnological conditions. Here we present results of analyses of a Lake Malawi core dating back ∼ 144 ka that describe and quantify the composition and abundance of fi sh fossils and report on stable carbon isotopic data ( δ 13
C) from fi sh scale, bone and tooth fossils. We compared the fossil δ 13
C values to δ 13
C values from extant fi sh communities to determine whether carbon isotope ratios can be used as indicators of inshore versus offshore pelagic fi sh assemblages. Fossil buccal teeth, pharyngeal teeth and mills, vertebra and scales from the fi sh families Cichlidae and Cyprinidae occur in variable abundance throughout the core.
Carbon isotopic ratios from numerous fi sh fossils throughout the core range between − 7.2 and − 27.5
‰ , similar to those found in contemporary Lake Malawi benthic and pelagic fi sh faunas. These results are the fi rst paleo-record of fi sh fossils from a Lake Malawi sediment core and the fi rst reported δ 13
C values from
Lake Malawi fi sh fossils. This approach provides a new methodology and framework for interpreting pelagic versus inshore fi sh faunas, lake level fl uctuations and the evolution of the Lake Malawi fi sh assemblages.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
In 2005 a series of drill cores was collected from two sites in Lake
Malawi/Nyasa, Africa ( Scholz et al., 2006, 2007
). Four cores were retrieved from a deep-water, central basin Site 1 (11°17.66S, 34°26.15E,
592 m water depth), one of which (MAL05-1C, 76 m in length) forms
the basis for this study. During other analyses of this core ( Cohen et al.,
2007
), it was noted by one of the authors (ASC) that numerous fi sh remains/fossils were present at all depths. These fossils included scales, bones and teeth. Whereas many types of chemical and biological indicators have been previously used to assess climate, lake level and faunal changes from Lake Malawi sediment cores, no one to date has quanti fi ed fossil fi sh remains or used these fossils as indicators of past fi sh communities or paleolimnological conditions. We examined these fossils in order to 1) determine which components are most commonly recovered from each species, 2) quantify their composition and abundance at varying levels of the core and 3) use stable carbon isotopic ratios to assess inshore versus offshore fi sh assemblages. Here we report the preliminary results of these analyses and discuss the applicability of using fi sh fossils as a methodology to examine lake level fl uctuations and evolutionary changes in an important African freshwater fi sh assemblage.
⁎ Corresponding author. Tel.: +1 520 621 7518; fax: +1 520 621 9190.
E-mail address: pnr@email.arizona.edu
(P.N. Reinthal).
0031-0182/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi: 10.1016/j.palaeo.2010.01.004
Lake Malawi contains one of the most diverse freshwater fi sh faunas found anywhere in the world (
Fryer and Iles, 1972; Ribbink et al., 1983 )
and is dominated primarily by members of the family Cichlidae.
Ecologically, the fi sh fauna may be segregated into those members of the inshore, rocky substrate fi sh assemblages versus the offshore pelagic and sandy bottom species. The near shore cichlids, known as ‘ mbuna ’ , are well known for their adaptive radiation and are considered a model system for evolutionary studies (
Kocher, 2004 ). Morphologically,
mbuna body forms are very similar but most cichlids have speciesspeci fi c oral (jaw or buccal) and pharyngeal dentition and neurocranial
morphology that are indicative of diet and habitat use ( Reinthal, 1990a, b
) (
Fig. 1 ). Alternatively, in the pelagic community of Lake Malawi,
four families are represented; Cichlidae, Cyprinidae, Clariidae and
Mochokidae. Only members from the fi rst two families demonstrate predominantly pelagic life history strategies. The pelagic cichlids are speci fi cally represented by three non-mbuna genera: Rhamphochromis
Regan, Diplotaxodon Trewavas and Copidichromis Eccles and Trewavas
(
Allison et al., 1995 ). The most abundant and widespread pelagic
fi sh species, and the only Lake Malawi species of fi sh to have pelagic planktonic larvae, is the cyprinid Engraulicypris sardella , a monospeci fi c genus of small neoboline cyprinid that spends its entire life in the pelagic zone of the lake (
Thompson, 1995 ).
Ecologically, the fi sh fauna of Lake Malawi is well known for its breadth of food resources. Stable isotopes (carbon δ 13 C and nitrogen
δ 15
N) have been used to determine food resource partitioning among

P.N. Reinthal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 303 (2011) 126 – 132 127
Fig. 1.
Variation in buccal (oral) teeth and neurocrania morphology among six species of Lake Malawi cichlids. All taxa shown here are members of the nearshore, monophyletic species complex known as ‘ mbuna ’ . Long, fl exible bicuspid and tricuspid teeth are found in algae grazers and rock scrappers. Shorter, unicuspid teeth are found in pelagic feeders
(drawing by D.S.C. Lewis).
Lake Malawi fi shes by comparing isotopes of a variety of inshore
benthic versus pelagic species and their available resources ( Bootsma et al., 1996
). Stable isotopes have also been used to look at niche
segregation among Lake Malawi cichlids ( Genner et al., 1999 ) and
ontogenetic variation in isotopic signatures related to body size
( Genner et al., 2003
). These studies provide a large sampling of extant inshore and pelagic fi sh isotopic signatures that may be used for comparative purposes. Species with δ 13 C values in the range of − 6 to
− 16 ‰ VPDB are primarily inshore mbuna species. Pelagic species, like Engraulicypris sardella , and open water cichlids show lower δ values, in the range of − 22 to − 25 ‰ VPDB.
13 C
Fish have long been known to play a critical role in the function and structure of lake ecosystems. It is only recently that scientists are beginning to assess the potential of fi sh fossils and stable isotope values from sediment cores in paleoecological studies (
Patterson and
Smith, 2002
). Scale and scale fragments have been used to examine percid – cyprinid faunal shifts in shallow English lakes (
Davidson et al.,
2003
). Stable isotopic values found in fi sh fossils have also been used in a variety of studies to evaluate paleolimnological conditions
( Patterson et al., 1993; Patterson, 1998; Smith and Patterson 1994;
Wurster and Patterson, 2001, 2003 ). For example, Nd and Sr isotopes
have been examined in Neogene fossil teeth to determine historical values of seawater isotopic values (
Martin and Haley, 2000 ), and in
archeological sites to determine the source of the fi
sh ( Dufour et al.,
2007
). Oxygen isotopes in otoliths have also been used to examine natal and adult temperature (
Dufour et al., 2005; Zazzo et al., 2006
), salinity and habitat preferences in Cynoscion othonopterus (
Rowell et al., 2005, 2008
), alewife ( Dufour et al., 2005; Dufour et al., 2008
) and undescribed Jurassic species (
Patterson, 1999 ). Lake sediment
records of δ 15 N and biological indicators were used to reconstruct salmon abundance in Alaska over the past 300 (
Finney et al., 2000
) and ∼ 2200 years (
Finney et al., 2002
).
Preliminary examination of sediment from the Lake Malawi core revealed a wide variety of well preserved fi sh teeth, bones and scales from all depths. We hypothesized that δ 13 C values of fi sh fossils would vary with fl uctuations in lake level.
‘ Mbuna ’ fossils and/or fossils with more positive δ 13 C values ( − 6 to − 16 ‰ VPDB) might be indicative of an inshore, shallow water fauna during periods of decreased or low lake levels. Alternatively, fossils of pelagic species with more negative
δ 13 C values ( − 22 to − 25 ‰ ) might be indicative of a deep-water fauna and higher lake levels.
The present study attempted to assess the paleolimnological potential of fi sh fossils from Lake Malawi cores by quantifying fossils and comparing fossils with contemporary structures and isotopic values. Speci fi cally, we sought to quantitatively examine teeth, bones and scale fi sh fossils found at different depths of the core, to identify modern analogs of these fossils, to identify fi sh assemblages based on fossils found in the cores and to use stable isotopes ( δ 13 C) to determine benthic versus pelagic food sources. We analyzed 467 stratigraphic horizons for fi sh fossil abundance (#/g) from throughout the core. We then collected δ 13 C data from fi sh fossils from selected intervals in the core. Finally, based on independent data sets from the same core we then assessed the potential of fi sh fossils as predictors of lake level, fi sh communities and climatic conditions.

128
2. Materials and methods
P.N. Reinthal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 303 (2011) 126 – 132
Core collection, geochronology and archiving methods are described elsewhere (
Scholz et al., 2006, 2007 ). MAL05-1C provides a
record of Lake Malawi history from ∼ 144 to 10 ka. Because of core over-penetration the most recent ∼ 10 ka are missing from the core record. Malawi core MAL05-1C was sampled at 16-cm intervals for a total of 467 samples collected for wet-sieved residue analysis. For a discussion of wet-sieved residue analysis see
Cohen et al. (2007)
.
Numbers of bone, scale and teeth fossils were counted from samples that had been disaggregated in deionized water and sieved using a
125-um stainless steel sieve. Samples were sorted using an Olympus
SZH stereomicroscope and separated for identi fi cation and isotopic analysis.
Stable isotope data was collected on a continuousfl ow gas-ratio mass spectrometer (Finnigan Delta PlusXL). Samples were combusted using an elemental analyzer (Costech) coupled to the mass spectrometer. Standardization is based on an acetanilide working standard calibrated against NBS-22 and USGS-24 for δ 13 C. A urea working standard was also used when processing extremely small samples.
Precision is normally better than ±0.1 for δ 13 C (1 σ ), based on repeated internal standards. Samples that were extremely small
(marked with an asterisk in
Table 1 ) have a higher uncertainty;
standards of equivalent size had a variability of ±0.25
‰ (1 σ ). Bone
Table 1
Core section and depth (cm) within core section (included for sample identi fi cation), fossil descriptions, weights of sample or combined samples when available, estimated model age of sample (from
Scholz et al., 2007
), and δ 13 C value. Carbon isotope ratios followed by an asterisk have an uncertainty of ± 0.25
‰ (1 σ ) due to small sample size.
Core section
1 H 1
8 H 2
8 H 2
8 H 2
8 H 2
8 H 2
9 H 1
9 H 1
9 H 1
10 H 3
11 H 1
11 H 1
11 H 1
11 H 1
11 H 1
12 H 2
12 H 2
14 H 1
14 H 3
14 H 3
15 H 2
17 H 1
17 H 1
17 H 2
20 E 3
1 H 1
1 H 2
1 H 2
1 H 2
1 H 2
1 H 2
4 H 3
1 H 1
1 H 1
1 H 1
1 H 1
1 H 1
1 H 1
1 H 1
Depth
(cm)
Fossil description
2 – 3 Unicuspid pharyngeal teeth
2 – 3
18 – 19
50 – 51
82 – 83
82 – 83
82 – 83
82 – 83
82 – 83
Four bone fragments
Three Cyprinid bones
Five bone fragments
One scale
Three teeth
Bone fragments
Bone fragments
Bone
41 – 42
41 – 42
41 – 42
Ten bone fragments
Bone fragments
Bone fragments
89 – 90
82 – 83
Four bone fragments
Five bone fragments
51.6
– 52.6
Five bone fragments
15.3
– 16.3
Two bones
47.3
– 48.3
Eleven bone fragments
111.4
– 112.4 Five vertebrae
111.4
– 112.4 Bone
111.4
– 112.4 Bone fragments
2.3
– 3.3
Eight vertebrae
18.3
– 19.3
Three irregular bones
50.3
– 51.3
Bone fragments
76.9
– 77.9
Bone fragments
66 – 67 Six scale fragments
66 – 67
82 – 83
82 – 83
98 – 99
Ten scales
Five scales
Bone Fragments
Bone Fragments
85.2
– 86.2
Six bone fragments
85.2
– 86.2
Three vertebrae
97.4
– 98.4
Three vertebrae
17.9
– 18.9
Eight scale fragments
17.9
– 18.9
Nine scale fragments
21.9
– 22.9
Five scales fragments
53.6
– 54.6
Oral tooth
53.6
– 54.6
Bone
30 – 31 Two pharyngeal teeth
43 – 44 Bone
Weight
(mg)
.047
Model age
(ybp)
10,283
− 21.63
− 23.31
− 8.59
− 11.95
− 7.16
− 22.12*
− 18.62*
− 18.64*
− 24.74*
− 24.17*
− 18.25*
− 21.66*
− 27.51
− 23.46*
− 24.55*
− 24.57
− 24.19
− 26.04*
− 23.90
− 22.71
− 22.76*
− 16.21
− 23.55
− 23.59*
− 23.75
− 21.29
− 20.96
− 25.10*
− 24.22*
− 25.17
− 21.41
− 23.73*
− 24.70*
− 23.67
− 24.35*
− 24.05
− 23.44*
− 23.50*
.001
.335
.016
.030
.037
.253
.822
.023
.314
.347
.129
.307
.135
.070
2.14
.283
.276
.048
.011
.008
.034
.208
.044
.164
.013
.058
.031
.011
.024
.049
.041
.012
.009
.025
.111
.019
.029
.133
61,576
62,011
62,884
62,884
62,884
63,500
63,718
64,145
71,043
72,513
72,513
72,731
72,731
72,949
78,519
78,519
85,181
86,952
86,952
88,301
96,812
96,812
97,256
111,617
10,283
10,662
11,433
12,222
12,222
12,222
12,222
12,222
13,444
13,444
13,444
14,692
14,692
39,224
Final δ 13
C value
− 20.29
and tooth samples were acidi fi ed with sulfurous acid (H
2
SO
3
) in silver foil capsules to remove structural carbonate from the bone apatite.
Two drops of acid were used followed by overnight drying in a 60 °C oven. This procedure was repeated once. Because the amount of organic carbon in these samples was extremely small (pre-acidi fi ed weights ranged from .001 to 2.14 mg with most being .05 to .2 mg), extra care was taken to eliminate carbon contamination. Silver foil capsules were heated to ∼ 500 °C in air to remove any carbon material on the foil.
3. Results
Identi fi able fi sh fossils of buccal teeth, pharyngeal teeth and mill, vertebrae, other bones and scales are all from the fi sh families
Cichlidae and Cyprinidae. All fi sh fossils are extremely small (less than one mm) and are found throughout the core with variation in abundance. More fossils were found in older sections than in younger sections (
Fig. 2 ). In sections of the core between 10 ka (youngest) and
60 ka, only 34.9% (44 of 126) of the sample horizons contained fi sh fossils, whereas in the older sections of the core (60 – ∼ 140 ka) this proportion rises to 78.7% (270 of 343 samples). Fish bones were the most common fossils, being present in 96.2% of samples with fossils present whereas fi sh scales were found in 36.0% and fi sh teeth in
24.2% of samples with fossils present respectively. Fish bones were also the most numerically abundant fossils found in the samples; when any fossils (bones, teeth or scales) were present in a sample
(314 samples), fi sh bones accounted for 82.8% of all of the fossils present.
Although it was impossible to taxonomically identify all of the bones to family and/or species, the cyprinid bones appeared to be dominated by, if not exclusively consist of ‘ Usipa ’ , Engraulicypris sardella , the pelagic cyprinid found in modern Lake Malawi. Given the taxonomic complexities of and limited morphological differentiation among the Lake Malawi cichlids, the cichlid bones were impossible to identify beyond family. However there were certain structures, like buccal (oral) teeth that are unmistakably from the family Cichlidae and most likely, from the mbuna group.
All of the bones were extremely small (less than 1 mm in length) and many of the bones were fragmented and therefore dif fi cult to identify (
Fig. 4
A – B). This also complicated our isotopic analyses. Most samples were too small to have enough material from single bones for isotopic analyses, requiring us to use multiple bones and bone fragments from the same sample to obtain enough organic material for carbon isotope analyses. We were able to obtain isotopic ratios from scale samples as small as .008 mg. We were not able to obtain isotopes from small tooth (.007 mg and .003 mg) or small bone (.001
and .006 mg) samples. Taxonomic speci fi c identi fi cations of bones and teeth to species were possible in a few cases and we were able to identify bone structures including abdominal and caudal vertebrae
(
Figs. 3
A and B). These are from the cyprinid, Engraulicypris sardella, and in two cases we found fused preural and ural centra (compound centrum) with urodermal and hypural plates diagnostic of cyprinids.
We also identi fi ed upper pharyngeal tooth pads and pharyngeal teeth (
Fig. 4 A). Pharyngeal teeth are found in both cichlids and
cyprinids, but most of these appeared to be cyprinids. Most of teeth found in the core were identi fi ed as pharyngeal teeth but we also
found oral (buccal) teeth from cichlids ( Fig. 5 A
– B). Cyprinids have no teeth in their oral jaws. The tricuspid oral teeth found in the core
(
Fig. 5
A – 5B) are from cichlids as they are found only in cichlid adult fi shes from modern Lake Malawi and no other fi shes. In adult cichlids, tricuspid oral teeth are found in certain species of algal scrapers in the mbuna (inshore, benthic) group but little is known about tooth shape development. Pharyngeal teeth were common and distinctive from
oral teeth ( Fig. 5 C
– D).
Numerous scale and scale fragments were found and they appeared to primarily cycloid scales. Cyprinids have exclusively

P.N. Reinthal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 303 (2011) 126 – 132
Fig. 2.
Log
10 concentration per gram of the summation the number of bone, scale and teeth fragments plus one. Data set starts at ∼ 10,000 BP due to core over-penetration.
cycloid scales. Cichlids are thought to have ctenoid scales, but one of us (PNR) has observed numerous cichlids, especially small juveniles, with cycloid scales.
We obtained a broad range of δ 13 C isotopic signatures from numerous fi sh fossils, bone organic matter, teeth, and scales, found in
various sections throughout the core ( Table 1 ,
Fig. 6 ). Stable carbon
isotopic ratios range between − 7.2 and − 27.5
‰ VPDB. Most of the δ 13 C values (34 of 41) were between − 20 and − 25 ‰ VPDB, indicative of a pelagic fauna. Seven samples had δ 13
C values that fell between − 7 and
− 19 ‰ VPDB, more typical of benthic inshore ‘ mbuna ’ cichlids. Multiple samples from within a single stratigraphic horizon (e.g. samples from
1H1 (2 – 3 cm), model age ∼ 10.3 ka; 1H1 (82 – 83 cm), model age
∼ 12.2 ka; 1H2 (41 – 42 cm), model age ∼ 13.4 ka; 8H2 (111.4
–
112.4 cm), model age ∼ 62.8 ka; 9H1 (50.3
– 51.3 cm), model age
∼ 64.1 ka; 11H1 (66 – 67 cm), model age ∼ 72.5 ka; 12H2 (85.2
–
86.2 cm), model age ∼ 78.5 ka; 14H3 (17.9
– 18.9 cm), model age ∼ 86.9 ka; 17H1 (53.6
– 54.6 cm), model age ∼ 96.8 ka) displayed variable δ 13
C values throughout the core ( Table 1
).
129
4. Discussion
The preliminary analyses presented here revealed that fi sh fossils are widespread throughout a sediment core from Lake Malawi and may provide valuable information regarding fi sh faunal composition and lake level fl uctuations. In core sections between 10 – 60 ka, when the MAL-05-1C core site in Lake Malawi is thought to have been about the same depth as at present ( ∼ 580 m), fi sh fossils were present in less than half as many horizons as was the case from samples ranging between 60 and 135 ka, when the lake was often much shallower
(and more broadly fl uctuating in depth). This may re fl ect greater fi sh productivity at the core site during periods of lower lake levels, when the core site would have been located closer to shore.
We were able to identify a number of different structures; teeth, bones and scales, from various taxonomic groups. This is the fi rst time such fossils have been identi fi ed and quanti fi ed from a Lake Malawi core. Unfortunately, many specimens are very small with limited diagnostic characteristics but we are able to identify some taxonomic groups. The results here are a preliminary analysis and it is expected that more complete quanti fi cation and identi fi cation will result in better resolution of faunal composition over the recent history of Lake Malawi.
However, even the data presented here provide insight into the evolutionary history of an important group of endemic cichlid fi shes often cited as examples of rapid and ‘ explosive ’ speciation
Kocher, 2004 .
For example, tricuspid teeth that are approximately 130,000 years old
( Fig. 5
A – B) from ‘ mbuna ’ inshore benthic species, indicates that this lineage is at least this age. Understanding dates of cichlid evolutionary events has proven problematic and attempts have focused on
extrapolating rates of gene evolution ( Meyer et al., 1990
), lake level fl
uctuation ( Sturmbauer et al., 2001; Johnson et al., 2002; Verheyen et al., 2003 ) and employing calibration from cichlid fossils ( Genner et al.,
2007 ). The fossils from the core can be used as calibration points for
estimating the timing of genetic divergence within the Lake Malawi cichlid group. Also, the minimum lake level for Lake Malawi for the past
140,000 years is hypothesized to have occurred approximately
108,500 ybp ( Cohen et al., 2007 ). These 130,000 year old
‘ mbuna ’ fossils provide support that the lineage was present prior to this major period of
Fig. 3.
A. Left lateral view of cyprinid ( Engraulicypris sardella ) preural centrum vertebrae from sample 5-H-2 (93.4
– 94.4 cm), 20.19 mblf (meters below lake fl oor), approximately
48,000 years old. Legend: ha — hemal arch, haz — hemal prezygapophysis, hpz — hemal postzygapophysis, hs — hemal spine, na — neural arch, naz — neural prezygapophysis, npz — neural postzygapophysis, ns — neural spine, pc — preural centrum.
Fig. 3 B. Three post Weberian vertebrae from sample 5-H-2 (93.4
– 94.4 cm), 20.19 mblf, approximately
48,000 years old.

130 P.N. Reinthal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 303 (2011) 126 – 132
Fig. 4.
A. View of cyprinid ( Engraulicypris sardella ) left upper pharyngeal plate and teeth (630×) from sample 1-H-1 (82 – 83 cm), 7.33 mblf (meters below lake fl oor), approximately
12,000 years old.
Fig. 4 B. Series of bone fragments from sample 12-H-2 (85.2
– 86.2 cm), 41.02 mblf, approximately 79,000 years old (125×).
desiccation and that at least some members of the lineage persisted through this major drying period. The presence of the ‘ mbuna ’ lineage prior to the 108.5 ka megadrought is consistent with nuclear (0.7 ma;
Won et al., 2006
2001
) and mitochondrial (0.57
– 1.0 ma;
Sturmbauer et al.,
) molecular clock estimates for mbuna and non-mbuna cichlid
divergence. However, a lake level fall of at least 600 m for Lake Malawi during the megadrought ( ∼ 108,500 ybp) would result in a major reduction in rocky habitat for cichlids with a greatly reduced region of the west coastline available as ‘ mbuna ’
habitat ( Cohen et al., 2007 ). While
there may have been signi fi cant reduction in the ‘ mbuna ’ diversity, various ‘ mbuna ’ lineages persisted and have undergone a subsequent species-level radiation as Lake Malawi reached its modern lake levels.
These fossils can provide an important resource for future complementary analyses. We were able to obtain δ 13 C isotopic ratios from extremely small fossils and such fossils may also provide raw material for DNA studies, tooth wear analyses and/or other biogeochemical studies.
Fig. 5.
A — Scanning electron microscope (SEM) image of cichlid oral (buccal) tricuspid tooth from sample 24-E-3 (58.1
– 59.1 cm), 78.67 mblf, approximately 444 µm in length and
130,000 years old.
Fig. 5
B — SEM image of oral tricuspid tooth from sample 24-E-3 (58.1
– 59.1 cm), 78.67 mblf (meters below lake fl oor), approximately 766 µm in length and 130,000 years old.
Fig. 5
C. SEM image of pharyngeal tooth from cichlid or cyprinid from sample 24-E-3 (58.1
– 59.1 cm), 78.67 mblf, which is approximately 826 µm in length and 130,000 years old.
Fig. 5
D. Light microscopy image of pharyngeal tooth from sample 24-E-3 (58.1
– 59.1 cm), 78.67 mblf, approximately 600 µm in length and 130,000 years old.

P.N. Reinthal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 303 (2011) 126 – 132 131
The lack of more benthic isotopic signatures (higher δ 13 C values) could also be the result of ontogenetic dietary shifts in cichlids.
Juvenile cichlids are found to have a mainly phytoplankton-based diet while sub-adults and adults consumed more epilithic algae. The mean
δ 13 C values of fi ve ontogenetic classes of Lake Malawi cichlids changed
considerably with increasing body size ( Genner et al., 2003 ). The
fossils used in this study were all extremely small and would be from juvenile cichlids.
δ 13 C values of the juvenile cichlids feeding on phytoplankton would be dif fi cult to distinguish from δ 13
C values from juvenile cyprinids feeding on a similar diet.
This is the fi rst paleo-record of fi sh fossils from Lake Malawi and the fi rst reported examples of δ 13
C values from Lake Malawi fi sh fossils. Such data are not only useful in understanding Lake Malawi's paleo-ichthyofauna, lake level fl uctuations and the evolution of cichlid fi shes but also provides a new methodology and framework for interpreting pelagic versus inshore fi sh faunas and climate change in the region.
Fig. 6.
δ 13
C values for fi sh fossils from Lake Malawi core (values on upper x axis and each value ( n = 41) represented by ● ) and principal component 1 scores (Stone. pers.
com) (values on lower x axis and represented by
―
) versus age (ypb) of samples based on the age model from
Scholz et al. (2007)
. The principal components analysis
(PCA) included the relative abundance of major diatom groups (see Stone et al., this volume, for further details), relative abundance of major ostracode groups, percentages of vivianite, terrigenous minerals, and carbonate-coated grains, and log-concentrations of chaoborid fragments and ostracode valves. PCA of 18 variables and 438 samples was performed using CANOCO 4.5 (
ter Braak and Š milauer, 2002
); variables were centered and standardized but otherwise untransformed. PCA scores were re-sampled to 200-yr intervals and smoothed by singular spectrum analysis, using AnalySeries 1.2 (
Paillard et al., 1996
). Positive PC1 scores represent minimum lake levels and negative PC1 scores represent maximum lake levels.
δ 13 C values indicating a benthic inshore fi sh were not found until Lake Malawi started undergoing signi fi cant lake level fl uctuations older than 60,000 ybp.
Acknowledgements
We thank reviewers William Patterson, John Lundberg, and Chris
Scholz for comments and suggestions that led to a greatly improved manuscript. Initial core processing was carried out at LacCore, the
National Lake Core Repository at the University of Minnesota. Funding for the fi eld program was provided by the U.S. National Science
Foundation – ESH Program and the International Continental Scienti fi c
Drilling Program, with additional funding for AC from the Smithsonian
Institution (ETE and Fellowship Programs). We also thank D. Gauggler,
R. Markus and L. Brooke McDonough for help with sample preparation.
This work was supported by NSF EAR-0602350 to the three authors.
We thank the government of Malawi for permission to undertake this research and Dr. D.S.C. Lewis for drawing
Fig. 1
.
The range of δ 13 C isotopic values from the organic matter in Lake
Malawi fi sh fossils is similar to that found in the contemporary fi sh fauna. Those species feeding primarily on plankton have low δ 13 C values and those species feeding on benthos have higher δ 13
The δ 13 C isotopic ratios from fi
C values.
sh fossils throughout the core show a range from − 7.2 to − 27.5
‰ VPDB, the same range as that found in tissue samples from contemporary Lake Malawi benthic and pelagic fi sh faunas (
Bootsma et al., 1996; Genner et al., 1999, 2003
). Most of the δ 13 C isotopic ratios found here are representative of a pelagic ecosystem ( δ 13
C ∼ − 22 to − 25 ‰ VPDB) and is dominated by cyprinids. Further targeted sampling is needed, especially of structures thought to be indicative of those species that primarily utilize benthic resources.
Numerous other factors, such as time averaging, ontogeny, metabolic processes, productivity, changes in the terrestrial environment, etc., can directly affect the isotopic signature of an organism
( Wurster and Patterson, 2003; O'Reilly et al., 2004; Wurster et al.,
2005; Diefendorf et al., 2007; Nonogaki et al., 2007 ) so caution is
advised in interpreting the δ 13 C isotopic ratios presented here. For example, the δ 13
C values from all living fi sh studies have used muscle tissue or whole fi sh for sampling. Given that δ 13 C values can vary among tissues within an individual, contemporary sampling of bone, scale and teeth needs to be conducted for comparative purposes.
Most of the δ 13 C values indicate a pelagic fauna. This could be due to the fact that much of the sampling was conducted in samples between ∼ 10 and 60 ka when the lake was deep and pelagic fi sh species expected to be at the core site. It is only in samples older than
60 ka did we fi nd δ 13 C values between − 7 and − 20 ‰ VPDB indicative of an inshore, benthic fi
sh fauna ( Fig. 6 ). Based on other
paleo-indicators, it is during this time that Lake Malawi underwent severe lake level fl uctuation. More δ 13 C values from samples dating between 60 and 135 ka are needed.
References
Allison, E.H., Ngatunga, B.P., Thompson, A.B., 1995. Identi fi cation of the pelagic fi shes.
In: Menz, A. (Ed.), The Fishery Potential and Productivity of the Pelagic Zone of Lake
Malawi/Niassa. Natural Resources Institute, Chatham, UK, pp. 159 – 178.
Bootsma, H.A., Hecky, R.E., Hesslein, R.H., Turner, G., 1996. Food partitioning among Lake
Malawi nearshore fi shes as revealed by stable isotope analyses. Ecology 77 (4),
1286 – 1290.
Cohen, A.S., Stone, J.R., Beuning, K.R.M., Park, L.E., Reinthal, P.N., Dettman, D.L., Scholz, C.A.,
Johnson, T.C., King, J.W., Talbot, M.R., Brown, E.T., Ivory, S.J., 2007. Ecological consequences of early Late Pleistocene megadroughts in tropical Africa. PNAS 104 (42),
16422 – 16427.
Davidson, T.A., Sayer, C.D., Perrow, M.R., Tomlinson, M.L., 2003. Representation of fi sh communities by scale sub-fossils in shallow lakes: implications for inferring percid – cyprinid shifts. J. Paleolimnology 30, 441 – 449.
Diefendorf, A.F., Patterson, W.P., Holmden, C., Mullins, H.T., 2007. Carbon isotopes of marl and lake sediment organic matter re fl ect terrestrial landscape change during the late Glacial and early Holocene (16,800 to 5,540 cal yr B.P.): a multiproxy study of lacustrine sediments at Lough Inchiquin, western Ireland. J. Paleoliminol. 39 (1),
101 – 115.
Dufour, E., Patterson, W.P., Höök, T.O., Rutherford, E.S., 2005. Early life history of Lake
Michigan alewives ( Alosa pseudoharengus ) inferred from intra-otolith stable isotope ratios. Can. J. Fish. Aquat. Sci. 62, 2362 – 2370.
Dufour, E., Holmden, C., Van Neer, W., Zazzo, A., Patterson, W.P., Degryse, P., Keppens, E.,
2007. Oxygen and strontium isotopes as provenance indicators of fi sh at archaeological sites: a case study from Sagalassos, SW Turkey. J. Archaeol. Sci. 34 (8), 1226 – 1239.
Dufour, E., Höök, T.O., Patterson, W.P., Rutherford, E.S., 2008. High-resolution isotope analysis of young alewife otoliths: assessment of temporal resolution and reconstruction of habitat occupancy and thermal history. J. Fish Biol. 73 (10),
2434 – 2451.
Finney, B.P., Gregory-Eaves, I., Sweetman, J., Douglas, M.S.V., Smol, J.P., 2000. Impacts of climate change and fi shing on Paci fi c salmon abundance over the past 300 years.
Science 290, 795 – 799.
Finney, B.P., Gregory-Eaves, I., Douglas, M.S.V., Smol, J.P., 2002. Fisheries productivity in the northeastern Paci fi c Ocean over the past 2,200 years. Nature 416, 729 – 733.
Fryer, G., Iles, T.D., 1972. The Cichlid Fishes of the Great Lakes of Africa. Oliver and Boyd,
Edinburgh. 641 pp.
Genner, M.A., Turner, G.F., Baker, S., Hawkins, S.J., 1999. Niche segregation among Lake
Malawi cichlid fi shes? Evidence from stable isotopic signatures. Ecol. Lett. 2,
185 – 190.
Genner, M.A., Hawkins, S.J., Turner, G.F., 2003. Isotopic change throughout the life history of a Lake Malawi cichlid fi sh. J. Fish Biol. 62, 907 – 917.

132 P.N. Reinthal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 303 (2011) 126 – 132
Genner, M.A., Seehausen, O., Lunt, D.H., Joyce, D.A., Shaw, P.W., Carvalho, G.R., Turner,
G.F., 2007. Age of cichlids: new dates for ancient lake fi sh radiations. Mol. Biol.
Evol. 24 (5), 1269 – 1282.
Johnson, T.C., Brown, E.T., McManus, J., Barry, S., Barker, P., Gasse, F., 2002. Highresolution paleoclimate record spanning the past 25,000 years in southern East
Africa. Science 296, 113 – 132.
Kocher, T.D., 2004. Adaptive evolution and explosive speciation: the cichlid fi sh model.
Nat. Rev., Genet. 5, 288 – 298.
Martin, E.F., Haley, B.A., 2000. Fossil fi sh teeth as proxies for seawater Sr and Nd isotopes. Geochim. Cosmochim. Acta 64 (5), 835 – 847.
Meyer, A., Kocher, T.D., Basasibwaki, P., Wilson, A.C., 1990. Monophyletic origin of Lake
Victoria cichlid fi shes suggested by mitochondrial DNA sequences. Nature 347,
550 – 553.
Nonogaki, H., Nelson, J., Patterson, W.P., 2007. Dietary histories of herbivorous Loricariid cat fi shes: evidence from 13C values of otoliths. Environ. Biol. Fishes 78 (1), 13 – 21.
O'Reilly, C.M., Verburg, P., Hecky, R.E., Plisnier, P.D., Cohen, A.S., 2004. Food web dynamics in stable isotope ecology: time integration of different trophic levels. In:
Seuront, L., Sutton, P. (Eds.), Handbook of scaling methods in aquatic ecology:
Measurement, analysis, simulation. CRC Press, Boca Raton, Florida, pp. 125 – 134.
Paillard, D., Labeyrie, L., Yiou, P., 1996. Macintosh program performs time-series analysis. EoS Trans. AGU 77, 379.
Patterson, W.P., 1998. North American continental seasonality during the last millennium: high-resolution analysis of sagittal otoliths. Palaeogeogr. Palaeoclimatol. Palaeoecol.
138 (no. 1 – 4), 271 – 303.
Patterson, W.P., 1999. Oldest isotopically characterized fi sh otoliths provide insight to
Jurassic continental climate of Europe. Geology 27 (3), 199 – 202.
Patterson, W.P., Smith, G.R., 2002. In: Smol, J.P., Birks, H.J.B., Last, W.M. (Eds.), Uses of fi shes in analysis of paleolimnology in tracking environmental change using lake sediments: zoological indicators: Paleolimnology, Special, vol. 4, pp. 173 – 187. ch. 9.
Patterson, W.P., Smith, G.R., Lohmann, K.C., 1993. Continental paleothermometry and seasonality using the isotopic composition of aragonitic otoliths of freshwater fi shes. Monogr. AGU 78, 191.
Reinthal, P.N., 1990a. Morphological analyses of the neurocranium of a group of rockdwelling cichlid fi shes (Cichlidae: Perciformes) from Lake Malawi, Africa. Zool. J.
Linn. Soc. (Lond.) 98, 123 – 139.
Reinthal, P.N., 1990b. The feeding habits of a group of tropical herbivorous rockdwelling cichlid fi shes (Cichlidae: Perciformes) from Lake Malawi, Africa. Environ.
Biol. Fishes 27, 215 – 233.
Ribbink, A.J., Marsh, B.A., Marsh, A.C., Ribbink, A.C., Sharp, B.J., 1983. A preliminary survey of the cichlid fi shes of rocky habitats in Lake Malawi. S. Afr. J. Zool. 18,
149 – 310.
Rowell, K., Flessa, K.W., Dettman, D.L., 2005. The importance of Colorado River fl ow to nursery habitats of Gulf corvina ( Cynoscion othonopterus ). Can. J. Fish. Aquat. Sci. 62,
2874 – 2885.
Rowell, K., Flessa, K.W., Dettman, D.L., Román, M.J., Gerber, L.R., Findley, L.T., 2008.
Diverting the Colorado River leads to a dramatic life history shift in an endangered marine fi sh. Biol. Conserv. 141, 1138 – 1148.
Scholz, C.A., Cohen, A.S., Johnson, T.C., King, J.W., Moran, K., 2006. The 2005 Lake Malawi
Scienti fi c Drilling Project. Sci. Drill. 2, 17 – 19.
Scholz, C.A., Johnson, T.C., Cohen, A.C., King, J.W., Peck, J.A., Overpeck, J.T., Talbot, M.R.,
Brown, E.R., Kalindekafe, L., Amoako, P.Y.O., Lyons, R.P., Shanahan, T.M., Castañeda,
I.S., Heil, C.W., Forman, S.L., McHargue, L.R., Beuning, K.R., Gomez, J., Pierson, J.,
2007. East African megadroughts between 135 and 75 thousand years ago and bearings on early-modern human origins. Proc. Natl. Acad. Sci. U. S. A. 104,
16416 – 16421.
Smith, G.R., Patterson, W.P., 1994. Mio-pliocene seasonality on the Snake River plain: comparison of faunal and oxygen isotopic evidence. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 107 (3 – 4), 291 – 302.
Sturmbauer, C., Baric, S., Salzburger, W., Ruber, L., Verheyen, E., 2001. Lake level fl uctuations synchronize genetic divergences of cichlid fi shes in African lakes. Mol.
Biol. Evol. 18, 144 – 154.
ter Braak, C.J.F., Š milauer, P., 2002. Canoco Reference Manual and CanoDraw for
Windows User's Guide: Software for Canonical Community Ordination (version
4.5). Microcomputer Power, Ithaca, NY, USA.
Thompson, A.B., 1995. Eggs and larvae of Engraulicypris sardella . In: Menz, A. (Ed.), The
Fishery Potential and Productivity of the Pelagic Zone of Lake Malawi/Niassa.
Natural Resources Institute, Chatham, UK, pp. 179 – 200.
Verheyen, E., Salzburger, W., Snoeks, J., Meyer, A., 2003. Origin of the super fl ock of cichlid fi shes from Lake Victoria, East Africa. Science 300, 325 – 329.
Won, Y., Wang, Y., Sivasundar, A., Raincrow, J., Hey, J., 2006. Nuclear gene variation and molecular dating of the cichlid species fl ock of Lake Malawi. Mol. Biol. Evol. 23,
828 – 837.
Wurster, C.M., Patterson, W.P., 2001. Late Holocene climate change for the eastern interior United States climate: evidence from high-resolution sagittal otolith stable isotope ratios of oxygen. Palaeogeogr. Palaeoclimatol. Palaeoecol. 170 (no. 1 – 2),
81 – 100.
Wurster, C.M., Patterson, W.P., 2003. Secular variation in metabolism of freshwater drum ( Aplodinotus grunniens ): evidence from stable carbon isotope values.
Paleobiology 29 (4), 492 – 505.
Wurster, C.M., Patterson, W.P., Stewart, D.J., Stewart, T.J., Bowlby, J.N., 2005. Thermal histories, stress, and metabolic rates of chinook salmon in Lake Ontario: evidence from intra-otolith 18O and 13C values and energetics modeling. Can. J. Fish. Aquat.
Sci. 62, 700 – 713.
Zazzo, A., Smith, G.R., Patterson, W.P., Dufour, E., 2006. Life history reconstruction of modern and fossil sockeye salmon ( Oncorhynchus nerka ) by oxygen isotopic analysis of otoliths, vertebrae, and teeth: implication for paleoenvironmental reconstructions. Earth Planet. Sci. Lett. 249 (3 – 4), 200 – 215.