Mini-Summary

Mini-Summary•The history of the earth is divided into geological time periods

Mini-Summary•The history of the earth is divided into geological time periods

• These are defined by characteristic flora and fauna

Mini-Summary•The history of the earth is divided into geological time periods

• These are defined by characteristic flora and fauna

•Large-scale changes in biodiversity were triggered by slow and rapid environmental change

Pg

K-Pg

(KT)

Tr-J

P-Tr

Late

DO-S

Today

Lecture 5: Fossils

1. How fossilization works. Some examples of fossils.

2. Dating fossils.

3. What we can learn from fossils?

y . wurm {@} qmul . ac .uk

Geological context

Geological context

Three broad classes of rock:

Geological context

Three broad classes of rock:

•Sedimentary rocks: formed by particles (mineral or organic) gradually settling out of solution, then compacting to form rock

Geological context

Three broad classes of rock:

•Sedimentary rocks: formed by particles (mineral or organic) gradually settling out of solution, then compacting to form rock

•Igneous rocks: formed by the cooling of magma

Geological context

Three broad classes of rock:

•Sedimentary rocks: formed by particles (mineral or organic) gradually settling out of solution, then compacting to form rock

•Igneous rocks: formed by the cooling of magma

•Metamorphic rocks: modification of existing rocks under high pressure and heat

Fossils: only in sedimentary rocks (deposited on oceanic shorelines, lake beds, flood plains...)

Fossils: only in sedimentary rocks (deposited on oceanic shorelines, lake beds, flood plains...)

Weathering or erosion can expose the older layers

Fossilization

Fossilization

• Permineralization

Fossilization

• Permineralization

• “Natural cast” process

Fossilization

• Permineralization

• “Natural cast” process

•Fossilization is rare & only in sediment...

Fossilization

• Permineralization

• “Natural cast” process

•Fossilization is rare & only in sediment...

•Ancient material also occurs:

• in amber

•by mummification

• in ice

Fossil formation at Sterkfontein

Fossil formation at SterkfonteinLimestone deposits were laid down 2.5 billion years ago when the area was a shallow sea.

Fossil formation at SterkfonteinLimestone deposits were laid down 2.5 billion years ago when the area was a shallow sea.

Caves eventually form below the surface.

Fossil formation at SterkfonteinLimestone deposits were laid down 2.5 billion years ago when the area was a shallow sea.

Caves eventually form below the surface.

‘Pot holes’ form between the surface and the caves.

Fossil formation at SterkfonteinLimestone deposits were laid down 2.5 billion years ago when the area was a shallow sea.

Caves eventually form below the surface.

‘Pot holes’ form between the surface and the caves.

Debris, including animals, fall in!

Fossil formation at SterkfonteinLimestone deposits were laid down 2.5 billion years ago when the area was a shallow sea.

Caves eventually form below the surface.

‘Pot holes’ form between the surface and the caves.

Debris, including animals, fall in!

Compaction and cementing with water and limestone produces “Breccia”.

Fossil preservation

•Hard part like shells, bones and teeth are usually all that remain

•Soft tissues fossils are rare

Why are fossils rare?

Why are fossils rare?

•Fossils don’t form often:

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation::

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation::• arid deserts, deep water (with low O2)

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation::• arid deserts, deep water (with low O2)

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation::• arid deserts, deep water (with low O2)

•

Why are fossils rare?

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation:• arid deserts, deep water (with low O2), cold

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation:• arid deserts, deep water (with low O2), cold

•Fossils can be lost:

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation:• arid deserts, deep water (with low O2), cold

•Fossils can be lost:•mountains: lots of erosion

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation:• arid deserts, deep water (with low O2), cold

•Fossils can be lost:•mountains: lots of erosion•Metamorphosis and subduction of rocks destroys fossils

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation:• arid deserts, deep water (with low O2), cold

•Fossils can be lost:•mountains: lots of erosion•Metamorphosis and subduction of rocks destroys fossils

Why are fossils rare?

•Fossils don’t form often: •Predators, scavengers, insects consume corpses•Bacteria and fungi decompose remains•Even faster in tropics (acid soil, warm, humid...)

•Best locations for fossil formation:• arid deserts, deep water (with low O2), cold

•Fossils can be lost:•mountains: lots of erosion•Metamorphosis and subduction of rocks destroys fossils

•Most are still buried rather than exposed at the surface

A few examples...

Aquatic reptile; not a dinosaur. But same time (Mesozoic Era).A typical fossil skeleton.

Plesiosaur fossil

More typical…

Parts of head, and anvil/brush of Akmonistion zangerli, shark from Carboniferous of Scotland

Ammonites

AmmoniteNautilus

Feathers, like soft tissue, are rarely preserved. But here imprinted in the rock.

Archaeopteryx - late Jurassic (150Mya)

“Fuzzy Raptor” (a dromaeosaur)

A rare example of a feathered dinosaur

The earliest Eutherian Mammal?Lower Cretaceous of China, 125 Mya

Eomaia scansoria

Ji et al., (2002) Nature 416, 816-822

A climbing mammal from a lake shore environment

Leptictidium tobieniEocene (Messel Shales, Germany)Soft tissues AND gut contents are

preserved

Dinosaur footprint

•At the time, this footprint of a dinosaur pressed into soft mud and became preserved in the now hardened rock. Can inform us on locomotion.

Fossilized tracks at Laetoli (Tanzania)

Footprints preserved in volcanic ash from: 3 hominids (Australopithecus afarensis)Numerous other mammals

Fossilized tracks at Laetoli (Tanzania)

Footprints preserved in volcanic ash from: 3 hominids (Australopithecus afarensis)Numerous other mammals

Fossil Ichthyosaur giving birth

•Such special preservations can inform us about the reproductive pattern in this species (live birth) .

Fossil Eggs

Information on development and social/reproductive behavior

INSECT IN AMBER

• This mosquito was imbedded in tree sap that subsequently hardened into amber, preserving the insect within.

Neanderthal skull from Iraq

(≈50,000 years old)

Very rarely, DNA can be extracted and

sequenced from such sub-fossils

Some animals get trapped in ice

Lecture 5: Fossils

1. How fossilization happens & some examples.

2. Dating fossils

3. What we can learn from fossils?

Dating methods

• Absolute - the item itself is dated

• Relative - strata above (younger) and below (older) are dated and the item expressed relative to these

Best method depends on context & age.

Principles of radiometric dating

Dating methods

Stratigraphy

Stratigraphy

As sediment collects, deeper layers are compacted by the ones above until they harden and become rock.Deeper Fossils are older than those above. Thus positions within the rock layers gives fossils a chronological age.

Index (Zone) Fossils

•Here, Locality 3 has no layer B (wasn’t formed or eroded).

•Index fossils: diagnostic fossil species that help dating new finds

Lecture 5: Fossils

1. How fossilization happens & some examples.

2. Dating fossils

3. What we can learn from fossils?

•Fossils can sometimes directly or indirectly tell us a great deal about the behavior of an organism, or its lifestyle

Intepreting fossils

•Careful interpretation: helps make sense of fossilized remains

•Analysis of hard parts can tell something about soft anatomy (e.g where muscles are (.e.g muscle scars).

•Geology: --> environmant (freshwater/marine/swamp))

• Infer from living organisms & relatives.

Hallucigenia sparsa (Cambrian Period)

From the Burgess Shale (Canada). Example of a soft bodied animal fossil, also very old!

Now re-interpreted as an Onychophoran ("velvet worm")

Colors don’t fossilize...

…or do they? (discovered fossilised melanosomes)

Colors don’t fossilize...

Fossils - Summary

Fossils - Summary

• Fossils form in sedimentary rock

• Fossilization is a rare process

• Usually, only the hard parts like bone, teeth, exoskeletons and shells are preserved

• Fossils of different ages occur in different strata, and “index fossils” can be used to cross-reference between different geographic locations

•Careful interpretation is required.

Lecture 6: What can DNA tell us?

Lecture 6: What can DNA tell us?•Species relationships previously based on:

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development•behavior

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development•behavior•ecological niche

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development•behavior•ecological niche•....

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development•behavior•ecological niche•....

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development•behavior•ecological niche•....

•DNA holds lots of additional information:

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development•behavior•ecological niche•....

•DNA holds lots of additional information:•2. Evolutionary relationships

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development•behavior•ecological niche•....

•DNA holds lots of additional information:•2. Evolutionary relationships•3. Current evolutionary contexts

Lecture 6: What can DNA tell us?•Species relationships previously based on:

•bone structures•morphologies •development•behavior•ecological niche•....

•DNA holds lots of additional information:•2. Evolutionary relationships•3. Current evolutionary contexts

1. DNA sequences change

1. DNA sequences change

1. DNA sequences change

DNA mutations occur all the time.

1. DNA sequences change

DNA mutations occur all the time. Reasons:

1. DNA sequences change

DNA mutations occur all the time. Reasons:

•mistakes in DNA replication or recombination,

1. DNA sequences change

DNA mutations occur all the time. Reasons:

•mistakes in DNA replication or recombination,•mutagens (radiation, chemicals),

1. DNA sequences change

DNA mutations occur all the time. Reasons:

•mistakes in DNA replication or recombination,•mutagens (radiation, chemicals),• viruses

1. DNA sequences change

DNA mutations occur all the time. Reasons:

•mistakes in DNA replication or recombination,•mutagens (radiation, chemicals),• viruses• transposons

1. DNA sequences change

DNA mutations occur all the time. Reasons:

•mistakes in DNA replication or recombination,•mutagens (radiation, chemicals),• viruses• transposons

1. DNA sequences change

DNA mutations occur all the time. Reasons:

•mistakes in DNA replication or recombination,•mutagens (radiation, chemicals),• viruses• transposons

Inherited: only if in germ line. Not from soma.

Types of mutations

Types of mutations•Small: replacement, insertion, deletion. E.g.:

Types of mutations•Small: replacement, insertion, deletion. E.g.:

original: GATTACAGATTACAnew : GATTACATATTACAPoint mutation

Types of mutations•Small: replacement, insertion, deletion. E.g.:

original: TGCAGATAGAGAGAGAGAGAGAGCAGATnew : TGCAGATAGAGAGAGAGAGCAGAT

Polymerase slippage in satellite

original: GATTACAGATTACAnew : GATTACATATTACAPoint mutation

Types of mutations•Small: replacement, insertion, deletion. E.g.:

•Big: inversions, duplications, deletions

original: TGCAGATAGAGAGAGAGAGAGAGCAGATnew : TGCAGATAGAGAGAGAGAGCAGAT

Polymerase slippage in satellite

original: GATTACAGATTACAnew : GATTACATATTACAPoint mutation

Types of mutations•Small: replacement, insertion, deletion. E.g.:

•Big: inversions, duplications, deletions

original: TGCAGATAGAGAGAGAGAGAGAGCAGATnew : TGCAGATAGAGAGAGAGAGCAGAT

Polymerase slippage in satellite

original: GATTACAGATTACAnew : GATTACATATTACAPoint mutation

Types of mutations•Small: replacement, insertion, deletion. E.g.:

•Big: inversions, duplications, deletions

original: TGCAGATAGAGAGAGAGAGAGAGCAGATnew : TGCAGATAGAGAGAGAGAGCAGAT

Polymerase slippage in satellite

original: GATTACAGATTACAnew : GATTACATATTACAPoint mutation

Types of mutations•Small: replacement, insertion, deletion. E.g.:

•Big: inversions, duplications, deletions

original: TGCAGATAGAGAGAGAGAGAGAGCAGATnew : TGCAGATAGAGAGAGAGAGCAGAT

Polymerase slippage in satellite

original: GATTACAGATTACAnew : GATTACATATTACAPoint mutation

Types of mutations•Small: replacement, insertion, deletion. E.g.:

•Big: inversions, duplications, deletions

original: TGCAGATAGAGAGAGAGAGAGAGCAGATnew : TGCAGATAGAGAGAGAGAGCAGAT

Polymerase slippage in satellite

original: GATTACAGATTACAnew : GATTACATATTACAPoint mutation

Mutations are the source of genetic, inheritable variation

What happens to a mutation?

What happens to a mutation?•Most point mutations are neutral: no effect.

What happens to a mutation?•Most point mutations are neutral: no effect.

• --> Genetic drift, hitchhiking... (--> elimination or fixation)

What happens to a mutation?•Most point mutations are neutral: no effect.

•Some are very deleterious;

• --> Genetic drift, hitchhiking... (--> elimination or fixation)

What happens to a mutation?•Most point mutations are neutral: no effect.

•Some are very deleterious;

• --> Genetic drift, hitchhiking... (--> elimination or fixation)

Eg. antennapedia (hox gene) mutation:

What happens to a mutation?•Most point mutations are neutral: no effect.

•Some are very deleterious;

• --> Genetic drift, hitchhiking... (--> elimination or fixation)

Some increase fitness.

Eg. antennapedia (hox gene) mutation:

What happens to a mutation?•Most point mutations are neutral: no effect.

•Some are very deleterious;

• --> selection eliminates or fixes them

•--> Genetic drift, hitchhiking... (--> elimination or fixation)

Some increase fitness.

Eg. antennapedia (hox gene) mutation:

What happens to a mutation?•Most point mutations are neutral: no effect.

•Some are very deleterious;

See population genetics lectures & practical

• --> selection eliminates or fixes them

•--> Genetic drift, hitchhiking... (--> elimination or fixation)

Some increase fitness.

Eg. antennapedia (hox gene) mutation:

2. DNA clarifies evolutionary relationships

2. DNA clarifies evolutionary relationships

Human: GATTACA

2. DNA clarifies evolutionary relationships

Human: GATTACAPeacock: GATTGCA

2. DNA clarifies evolutionary relationships

Human: GATTACAPeacock: GATTGCA Amoeba: GGCTCCA

2. DNA clarifies evolutionary relationships

Human: GATTACAPeacock: GATTGCA Amoeba: GGCTCCA

Human

Peacock

Amoeba

2. DNA clarifies evolutionary relationships

Human: GATTACAPeacock: GATTGCA Amoeba: GGCTCCA

Human

Peacock

Amoeba

See practical!

Linnaeus 1735 classification of animals

Carl Woese in 1977

Cows are more closey related to whales than to horses

Artiodactyla

Cetartiodactyla

Cetacea

Bat echolocation

Teeling 2002

Bat echolocation

Teeling 2002

Evolved twice!

Molecular clocks: another dating system

Stochastic clock

Molecular clocks: another dating system

Stochastic clock

Time

Gen

etic

cha

nge

Metronomicclock

Molecular clocks: another dating system

Stochastic clock

Time

Gen

etic

cha

nge

Metronomicclock

Mutations may accumulate with time at different rates

Ancient DNA

Ancient DNA: below 2km of ice

Ancient DNA: below 2km of icecriterion of authenticity obviously dismissesmany putative taxa that are present at lowabundance or have heterogeneous distributions,as is typical of environmental samples (16), butefficiently minimizes the influence of possiblelow-level contamination and misidentificationsdue to DNA damage (17).

Approximately 31% of the sequences fromthe John Evans Glacier silty sample were as-signed to plant taxa that passed the authentica-tion and identification criteria. These belong tothe order Rosales, the family Salicaceae, and thegenus Saxifraga (Table 1). This result is con-sistent with the John Evans Glacier forming nomore than a few thousand years ago in a highArctic environment (18), characterized by lowplant diversity and sparse vegetation cover sim-ilar to that currently surrounding the glacier,which consists mainly of Arctic willow (Salica-ceae), purple saxifrage (Saxifraga), Dryas (Ro-sales), and Arctic poppy (Papaver) (19). Thus,by confirming the expected result, the John EvansGlacier study can be regarded as a positive con-trol, showing that DNA data from silty ice reli-ably record the local ecology.

In contrast to the John Evans Glacier siltysample, the 45% of the Dye 3 DNA sequencesthat could be assigned to taxa reveal a commu-nity very different from that of Greenland today.The taxa identified include trees such as alder(Alnus), spruce (Picea), pine (Pinus), and mem-bers of the yew family (Taxaceae) (Table 1).Their presence indicates a northern boreal for-est ecosystem rather than today’s Arctic environ-ment. The other groups identified, includingAsteraceae, Fabaceae, and Poaceae, are mainly

Table 1. Plant and insect taxa obtained from the JEG and Dye 3 silty icesamples. For each taxon (assigned to order, family, or genus level), thegenetic markers (rbcL, trnL, or COI), the number of clone sequencessupporting the identification, and the probability support (in percentage)

are shown. Sequences have been deposited in GenBank under accessionnumbers EF588917 to EF588969, except for seven sequences less than 50bp in size that are shown below. Their taxon identifications are indicatedby symbols.

Order Marker Clones Support (%) Family Marker Clones Support (%) Genus Marker Clones Support (%)JEG sampleRosales rbcL 3 90–99Malpighiales rbcL

trnL25

99–10099–100

Salicaceae rbcLtrnL

24

99–100100

Saxifragales rbcL 3 92–94 Saxifragaceae rbcL 2 92 Saxifraga rbcL 2 91Dye 3 sampleConiferales rbcL

trnL4427

97–100100

Pinaceae* rbcLtrnL

2025

100100

PiceaPinus†

rbcLtrnL

2017

99–10090–99

Taxaceae‡ rbcLtrnL

232

91–98100

Poales§ rbcLtrnL

6717

99–10097–100

Poaceae§ rbcLtrnL

6713

99–100100

Asterales rbcLtrnL

1827

90–100100

Asteraceae rbcLtrnL

227

91100

Fabales rbcLtrnL

103

99–10099

Fabaceae rbcLtrnL

103

99–10099

Fagales rbcLtrnL

1012

95–99100

Betulaceae rbcLtrnL

811

93–9798–100

Alnus rbcLtrnL

79

91–9598–100

Lepidoptera COI 12 97–99*Env_2, trnL ATCCGGTTCATGAAGACAATGTTTCTTCTCCTAAGATAGGAAGGG. Env_3, trnL ATCCGGTTCATGAAGACAATGTTTCTTCTCCTAATATAGGAAGGG. Env_4, trnL ATCCGGTTCATGAGGACAATGTTTCTTCTCCTAATA-TAGGAAGGG. †Env_5, trnL CCCTTCCTATCTTAGGAGAAGAAACATTGTCTTCATGAACCGGAT. Env_6, trnL TTTCCTATCTTAGGAGAAGAAACATTGTCTTCATGAACCGGAT. ‡Env_1, trnL ATCCGTATTATAG-GAACAATAATTTTATTTTCTAGAAAAGG. §Env_7, trnL CTTTTCCTTTGTATTCTAGTTCGAGAATCCCTTCTCAAAACACGGAT.

Fig. 1. Sample location and core schematics. (A) Map showing the locations of the Dye 3 (65°11'N,45°50'W) and GRIP (72°34'N, 37°37'W) drilling sites and the Kap København Formation (82°22'N,W21°14'W) in Greenland as well as the John Evans Glacier (JEG) (79°49'N, 74°30'W) on EllesmereIsland (Canada). The inset shows the ratio of D– to L–aspartic acid, a measure of the extent of proteindegradation; more highly degraded samples (above the line) failed to yield amplifiable DNA. (B)Schematic drawing of ice core/icecap cross section, with depth [recorded in meters below thesurface (m.b.s.)] indicating the depth of the cores and the positions of the Dye 3, GRIP, and JEGsamples analyzed for DNA, DNA/amino acid racemization/luminescence (underlined), and 10Be/36Cl(italic). The control GRIP samples are not shown. The lengths (in meters) of the silty sections arealso shown.

6 JULY 2007 VOL 317 SCIENCE www.sciencemag.org112

REPORTS

on

Janu

ary

21, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fro

m

criterion of authenticity obviously dismissesmany putative taxa that are present at lowabundance or have heterogeneous distributions,as is typical of environmental samples (16), butefficiently minimizes the influence of possiblelow-level contamination and misidentificationsdue to DNA damage (17).

Approximately 31% of the sequences fromthe John Evans Glacier silty sample were as-signed to plant taxa that passed the authentica-tion and identification criteria. These belong tothe order Rosales, the family Salicaceae, and thegenus Saxifraga (Table 1). This result is con-sistent with the John Evans Glacier forming nomore than a few thousand years ago in a highArctic environment (18), characterized by lowplant diversity and sparse vegetation cover sim-ilar to that currently surrounding the glacier,which consists mainly of Arctic willow (Salica-ceae), purple saxifrage (Saxifraga), Dryas (Ro-sales), and Arctic poppy (Papaver) (19). Thus,by confirming the expected result, the John EvansGlacier study can be regarded as a positive con-trol, showing that DNA data from silty ice reli-ably record the local ecology.

In contrast to the John Evans Glacier siltysample, the 45% of the Dye 3 DNA sequencesthat could be assigned to taxa reveal a commu-nity very different from that of Greenland today.The taxa identified include trees such as alder(Alnus), spruce (Picea), pine (Pinus), and mem-bers of the yew family (Taxaceae) (Table 1).Their presence indicates a northern boreal for-est ecosystem rather than today’s Arctic environ-ment. The other groups identified, includingAsteraceae, Fabaceae, and Poaceae, are mainly

Table 1. Plant and insect taxa obtained from the JEG and Dye 3 silty icesamples. For each taxon (assigned to order, family, or genus level), thegenetic markers (rbcL, trnL, or COI), the number of clone sequencessupporting the identification, and the probability support (in percentage)

are shown. Sequences have been deposited in GenBank under accessionnumbers EF588917 to EF588969, except for seven sequences less than 50bp in size that are shown below. Their taxon identifications are indicatedby symbols.

Order Marker Clones Support (%) Family Marker Clones Support (%) Genus Marker Clones Support (%)JEG sampleRosales rbcL 3 90–99Malpighiales rbcL

trnL25

99–10099–100

Salicaceae rbcLtrnL

24

99–100100

Saxifragales rbcL 3 92–94 Saxifragaceae rbcL 2 92 Saxifraga rbcL 2 91Dye 3 sampleConiferales rbcL

trnL4427

97–100100

Pinaceae* rbcLtrnL

2025

100100

PiceaPinus†

rbcLtrnL

2017

99–10090–99

Taxaceae‡ rbcLtrnL

232

91–98100

Poales§ rbcLtrnL

6717

99–10097–100

Poaceae§ rbcLtrnL

6713

99–100100

Asterales rbcLtrnL

1827

90–100100

Asteraceae rbcLtrnL

227

91100

Fabales rbcLtrnL

103

99–10099

Fabaceae rbcLtrnL

103

99–10099

Fagales rbcLtrnL

1012

95–99100

Betulaceae rbcLtrnL

811

93–9798–100

Alnus rbcLtrnL

79

91–9598–100

Lepidoptera COI 12 97–99*Env_2, trnL ATCCGGTTCATGAAGACAATGTTTCTTCTCCTAAGATAGGAAGGG. Env_3, trnL ATCCGGTTCATGAAGACAATGTTTCTTCTCCTAATATAGGAAGGG. Env_4, trnL ATCCGGTTCATGAGGACAATGTTTCTTCTCCTAATA-TAGGAAGGG. †Env_5, trnL CCCTTCCTATCTTAGGAGAAGAAACATTGTCTTCATGAACCGGAT. Env_6, trnL TTTCCTATCTTAGGAGAAGAAACATTGTCTTCATGAACCGGAT. ‡Env_1, trnL ATCCGTATTATAG-GAACAATAATTTTATTTTCTAGAAAAGG. §Env_7, trnL CTTTTCCTTTGTATTCTAGTTCGAGAATCCCTTCTCAAAACACGGAT.

Fig. 1. Sample location and core schematics. (A) Map showing the locations of the Dye 3 (65°11'N,45°50'W) and GRIP (72°34'N, 37°37'W) drilling sites and the Kap København Formation (82°22'N,W21°14'W) in Greenland as well as the John Evans Glacier (JEG) (79°49'N, 74°30'W) on EllesmereIsland (Canada). The inset shows the ratio of D– to L–aspartic acid, a measure of the extent of proteindegradation; more highly degraded samples (above the line) failed to yield amplifiable DNA. (B)Schematic drawing of ice core/icecap cross section, with depth [recorded in meters below thesurface (m.b.s.)] indicating the depth of the cores and the positions of the Dye 3, GRIP, and JEGsamples analyzed for DNA, DNA/amino acid racemization/luminescence (underlined), and 10Be/36Cl(italic). The control GRIP samples are not shown. The lengths (in meters) of the silty sections arealso shown.

6 JULY 2007 VOL 317 SCIENCE www.sciencemag.org112

REPORTS

on

Janu

ary

21, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fro

m

Willerslev 2007

criterion of authenticity obviously dismissesmany putative taxa that are present at lowabundance or have heterogeneous distributions,as is typical of environmental samples (16), butefficiently minimizes the influence of possiblelow-level contamination and misidentificationsdue to DNA damage (17).

Approximately 31% of the sequences fromthe John Evans Glacier silty sample were as-signed to plant taxa that passed the authentica-tion and identification criteria. These belong tothe order Rosales, the family Salicaceae, and thegenus Saxifraga (Table 1). This result is con-sistent with the John Evans Glacier forming nomore than a few thousand years ago in a highArctic environment (18), characterized by lowplant diversity and sparse vegetation cover sim-ilar to that currently surrounding the glacier,which consists mainly of Arctic willow (Salica-ceae), purple saxifrage (Saxifraga), Dryas (Ro-sales), and Arctic poppy (Papaver) (19). Thus,by confirming the expected result, the John EvansGlacier study can be regarded as a positive con-trol, showing that DNA data from silty ice reli-ably record the local ecology.

In contrast to the John Evans Glacier siltysample, the 45% of the Dye 3 DNA sequencesthat could be assigned to taxa reveal a commu-nity very different from that of Greenland today.The taxa identified include trees such as alder(Alnus), spruce (Picea), pine (Pinus), and mem-bers of the yew family (Taxaceae) (Table 1).Their presence indicates a northern boreal for-est ecosystem rather than today’s Arctic environ-ment. The other groups identified, includingAsteraceae, Fabaceae, and Poaceae, are mainly

Table 1. Plant and insect taxa obtained from the JEG and Dye 3 silty icesamples. For each taxon (assigned to order, family, or genus level), thegenetic markers (rbcL, trnL, or COI), the number of clone sequencessupporting the identification, and the probability support (in percentage)

are shown. Sequences have been deposited in GenBank under accessionnumbers EF588917 to EF588969, except for seven sequences less than 50bp in size that are shown below. Their taxon identifications are indicatedby symbols.

Order Marker Clones Support (%) Family Marker Clones Support (%) Genus Marker Clones Support (%)JEG sampleRosales rbcL 3 90–99Malpighiales rbcL

trnL25

99–10099–100

Salicaceae rbcLtrnL

24

99–100100

Saxifragales rbcL 3 92–94 Saxifragaceae rbcL 2 92 Saxifraga rbcL 2 91Dye 3 sampleConiferales rbcL

trnL4427

97–100100

Pinaceae* rbcLtrnL

2025

100100

PiceaPinus†

rbcLtrnL

2017

99–10090–99

Taxaceae‡ rbcLtrnL

232

91–98100

Poales§ rbcLtrnL

6717

99–10097–100

Poaceae§ rbcLtrnL

6713

99–100100

Asterales rbcLtrnL

1827

90–100100

Asteraceae rbcLtrnL

227

91100

Fabales rbcLtrnL

103

99–10099

Fabaceae rbcLtrnL

103

99–10099

Fagales rbcLtrnL

1012

95–99100

Betulaceae rbcLtrnL

811

93–9798–100

Alnus rbcLtrnL

79

91–9598–100

Lepidoptera COI 12 97–99*Env_2, trnL ATCCGGTTCATGAAGACAATGTTTCTTCTCCTAAGATAGGAAGGG. Env_3, trnL ATCCGGTTCATGAAGACAATGTTTCTTCTCCTAATATAGGAAGGG. Env_4, trnL ATCCGGTTCATGAGGACAATGTTTCTTCTCCTAATA-TAGGAAGGG. †Env_5, trnL CCCTTCCTATCTTAGGAGAAGAAACATTGTCTTCATGAACCGGAT. Env_6, trnL TTTCCTATCTTAGGAGAAGAAACATTGTCTTCATGAACCGGAT. ‡Env_1, trnL ATCCGTATTATAG-GAACAATAATTTTATTTTCTAGAAAAGG. §Env_7, trnL CTTTTCCTTTGTATTCTAGTTCGAGAATCCCTTCTCAAAACACGGAT.

Fig. 1. Sample location and core schematics. (A) Map showing the locations of the Dye 3 (65°11'N,45°50'W) and GRIP (72°34'N, 37°37'W) drilling sites and the Kap København Formation (82°22'N,W21°14'W) in Greenland as well as the John Evans Glacier (JEG) (79°49'N, 74°30'W) on EllesmereIsland (Canada). The inset shows the ratio of D– to L–aspartic acid, a measure of the extent of proteindegradation; more highly degraded samples (above the line) failed to yield amplifiable DNA. (B)Schematic drawing of ice core/icecap cross section, with depth [recorded in meters below thesurface (m.b.s.)] indicating the depth of the cores and the positions of the Dye 3, GRIP, and JEGsamples analyzed for DNA, DNA/amino acid racemization/luminescence (underlined), and 10Be/36Cl(italic). The control GRIP samples are not shown. The lengths (in meters) of the silty sections arealso shown.

6 JULY 2007 VOL 317 SCIENCE www.sciencemag.org112

REPORTS

on

Janu

ary

21, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fro

m

23. L. Mehl, B. R. Hacker, G. Hirth, P. B. Kelemen,J. Geophys. Res. 108, 10.1029/2002JB002233(2003).

24. B. R. Jicha, D. W. Scholl, B. S. Singer, G. M. Yogodzinski,S. M. Kay, Geology 34, 661 (2006).

25. C.-T. Lee, X. Cheng, U. Horodyskyj, Contrib. Min. Petrol.151, 222 (2006).

26. P. B. Kelemen, K. Hanghøj, A. R. Greene, in The Crust,

R. L. Rudnick, Ed. (Elsevier-Pergamon, Oxford, 2003),vol. 3, pp. 593–659.

27. T. V. Gerya, D. A. Yuen, Earth Planet. Sci. Lett. 212, 47(2003).

28. We thank M. Long, E. Kneller, and C. Conrad forconversations that motivated this work. Funding wasprovided by NSF grants EAR-9910899, EAR-0125919, andEAR-0509882.

Supporting Online Materialwww.sciencemag.org/cgi/content/full/317/5834/108/DC1SOM TextFigs. S1 and S2References

13 February 2007; accepted 9 May 200710.1126/science.1141269

Ancient Biomolecules fromDeep Ice Cores Reveal a ForestedSouthern GreenlandEske Willerslev,1* Enrico Cappellini,2 Wouter Boomsma,3 Rasmus Nielsen,4Martin B. Hebsgaard,1 Tina B. Brand,1 Michael Hofreiter,5 Michael Bunce,6,7Hendrik N. Poinar,7 Dorthe Dahl-Jensen,8 Sigfus Johnsen,8 Jørgen Peder Steffensen,8Ole Bennike,9 Jean-Luc Schwenninger,10 Roger Nathan,10 Simon Armitage,11Cees-Jan de Hoog,12 Vasily Alfimov,13 Marcus Christl,13 Juerg Beer,14 Raimund Muscheler,15Joel Barker,16 Martin Sharp,16 Kirsty E. H. Penkman,2 James Haile,17 Pierre Taberlet,18M. Thomas P. Gilbert,1 Antonella Casoli,19 Elisa Campani,19 Matthew J. Collins2

It is difficult to obtain fossil data from the 10% of Earth’s terrestrial surface that is covered by thickglaciers and ice sheets, and hence, knowledge of the paleoenvironments of these regions hasremained limited. We show that DNA and amino acids from buried organisms can be recoveredfrom the basal sections of deep ice cores, enabling reconstructions of past flora and fauna. Weshow that high-altitude southern Greenland, currently lying below more than 2 kilometers of ice,was inhabited by a diverse array of conifer trees and insects within the past million years. Theresults provide direct evidence in support of a forested southern Greenland and suggest that manydeep ice cores may contain genetic records of paleoenvironments in their basal sections.

The environmental histories of high-latituderegions such as Greenland and Antarcticaare poorly understood because much of

the fossil evidence is hidden below kilometer-thick ice sheets (1–3). We test the idea that thebasal sections of deep ice cores can act asarchives for ancient biomolecules.

The samples studied come from the basalimpurity-rich (silty) ice sections of the 2-km-long Dye 3 core from south-central Greenland(4), the 3-km-long Greenland Ice Core Project(GRIP) core from the summit of the Greenlandice sheet (5), and the Late Holocene John EvansGlacier on Ellesmere Island, Nunavut, northernCanada (Fig. 1). The last-mentioned sample wasincluded as a control to test for potential exoticDNA because the glacier has recently overriddena land surface with a known vegetation cover(6). As an additional test for long-distanceatmospheric dispersal of DNA, we includedfive control samples of debris-free Holoceneand Pleistocene ice taken just above the basalsilty samples from the Dye 3 and GRIP icecores (Fig. 1B). Finally, our analyses includedsediment samples from the Kap KøbenhavnFormation from the northernmost part ofGreenland, dated to 2.4 million years beforethe present (Ma yr B.P.) (1, 2).

The silty ice yielded only a few pollen grainsand no macrofossils (7). However, the Dye 3and John Evans Glacier silty ice samples showedlow levels of amino acid racemization (Fig. 1A,inset), indicating good organic matter preserva-tion (8). Therefore, after previous success withpermafrost and cave sediments (9–11), we at-tempted to amplify ancient DNA from the ice.This was done following strict criteria to secureauthenticity (12–14), including covering the sur-

face of the frozen cores with plasmid DNA tocontrol for potential contamination that mayhave entered the interior of the samples throughcracks or during the sampling procedure (7).Polymerase chain reaction (PCR) products ofthe plasmid DNA were obtained only from ex-tracts of the outer ice scrapings but not from theinterior, confirming that sample contaminationhad not penetrated the cores.

Using PCR, we could reproducibly amplifyshort amplicons [59 to 120 base pairs (bp)] ofthe chloroplast DNA (cpDNA) rbcL gene andtrnL intron from ~50 g of the interior ice meltsfrom the Dye 3 and the John Evans Glacier siltysamples. From Dye 3, we also obtained 97-bpamplicons of invertebrate cytochrome oxidasesubunit I (COI) mitochondrial DNA (mtDNA).Attempts to reproducibly amplify DNA fromthe GRIP silty ice and from the Kap KøbenhavnFormation sediments were not successful. Theseresults are consistent with the amino acid race-mization data demonstrating superior preser-vation of biomolecules in the Dye 3 and JohnEvans Glacier silty samples, which is likelybecause these samples are colder (Dye 3) oryounger (John Evans Glacier) than the GRIPsample (Fig. 1A, inset). We also failed to amplifyDNA from the five control samples of Holoceneand Pleistocene ice taken just above the siltysamples from the Dye 3 and GRIP ice cores(volumes: 100 g to 4 kg; Fig. 1B) (7). None ofthe samples studied yielded putative sequencesof vertebrate mtDNA.

A previous study has shown that simple com-parisons of short DNA sequences to GenBanksequences by means of the Basic Local Align-ment Search Tool (BLAST) make misidentifica-tion likely (15). Therefore, we assigned theobtained sequences to the taxonomic levels oforder, family, or genus using a new rigorousstatistical approach (7). In brief, this Bayesianmethod calculates the probability that each se-quence belongs to a particular clade by consid-ering its position in a phylogenetic tree based onsimilar GenBank sequences. In the calculationof these probabilities, uncertainties regarding phy-logeny, models of evolution, and missing dataare taken into account. Sequences with >90%posterior probability of membership to a taxo-nomic group were assigned to that group. Addi-tionally, a given plant taxon was only consideredgenuine if sequences assigned to that taxon werefound to be reproducibly obtained in separateanalyses (by independent laboratories for theDye 3 sample and within the laboratory for theJohn Evans Glacier control sample). This strict

1Centre for Ancient Genetics, University of Copenhagen,Denmark. 2BioArch, Departments of Biology and Archaeology,University of York, UK. 3Bioinformatics Centre, University ofCopenhagen, Denmark. 4Centre for Comparative Genomics,University of Copenhagen, Denmark. 5Max Planck Institute forEvolutionary Anthropology, Germany. 6Murdoch UniversityAncient DNA Research Laboratory, Murdoch University,Australia. 7McMaster Ancient DNA Center, McMaster Uni-versity, Canada. 8Ice and Climate, University of Copenhagen,Denmark. 9Geological Survey of Denmark and Greenland,Denmark. 10Research Laboratory for Archaeology and theHistory of Art, University of Oxford, UK. 11Department ofGeography, Royal Holloway, University of London, UK.12Department of Earth Sciences, University of Oxford, UK.13Paul Scherrer Institut (PSI)/Eidgenössische TechnischeHochschule (ETH) Laboratory for Ion Beam Physics, Institutefor Particle Physics, ETH Zurich, Switzerland. 14Swiss FederalInstitute of Aquatic Science and Technology (EAWAG),Switzerland. 15GeoBiosphere Science Center, Lund Univer-sity, Sweden. 16Department of Earth and AtmosphericSciences, University of Alberta, Canada. 17Ancient Biomole-cules Centre, Oxford University, UK. 18Laboratoire d'EcologieAlpine, CNRS Unité Mixte de Recherche 5553, UniversitéJoseph Fourier, Boîte Postale 53, 38041 Grenoble Cedex 9,France. 19Dipartimento di Chimica Generale e Inorganica,Università di Parma, Italy.

*To whom correspondence should be addressed. E-mail:ewillerslev@bi.ku.dk

www.sciencemag.org SCIENCE VOL 317 6 JULY 2007 111

REPORTS

on

Janu

ary

21, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fro

m

Willerslev 2007

Ancient DNA

Ancient DNA

Hair sequencing

Consensus with fossil record?

Genomic Evidence for Large, Long-Lived Ancestors toPlacental MammalsJ. Romiguier,1 V. Ranwez,1,2 E.J.P. Douzery,1 and N. Galtier*,1

1CNRS, Universite Montpellier 2, UMR 5554, ISEM, Montpellier, France2Montpellier SupAgro, UMR 1334, AGAP, Montpellier, France

*Corresponding author: E-mail: nicolas.galtier@univ-montp2.fr.

Associate editor: Naruya Saitou

Abstract

It is widely assumed that our mammalian ancestors, which lived in the Cretaceous era, were tiny animals that survived massiveasteroid impacts in shelters and evolved into modern forms after dinosaurs went extinct, 65 Ma. The small size of most Mesozoicmammalian fossils essentially supports this view. Paleontology, however, is not conclusive regarding the ancestry of extantmammals, because Cretaceous and Paleocene fossils are not easily linked to modern lineages. Here, we use full-genome data toestimate the longevity and body mass of early placental mammals. Analyzing 36 fully sequenced mammalian genomes, wereconstruct two aspects of the ancestral genome dynamics, namely GC-content evolution and nonsynonymous over synonym-ous rate ratio. Linking these molecular evolutionary processes to life-history traits in modern species, we estimate that earlyplacental mammals had a life span above 25 years and a body mass above 1 kg. This is similar to current primates, cetartiodactyls,or carnivores, but markedly different from mice or shrews, challenging the dominant view about mammalian origin andevolution. Our results imply that long-lived mammals existed in the Cretaceous era and were the most successful in evolution,opening new perspectives about the conditions for survival to the Cretaceous–Tertiary crisis.

Key words: phylogeny, GC-content, dN/dS ratio, GC-biased gene conversion, placentalia, fossils.

IntroductionIt is commonly assumed that early mammals were small crea-tures that only evolved into a variety of forms and sizes afterthe massive extinction of large reptiles, at the Cretaceous/Tertiary (KT) boundary, 65 Ma (Dawkins 2004; Feldhameret al. 2007). This scenario is consistent with theoretical con-siderations: Cope’s rule (Alroy 1998) states that current livinglineages generally descend from small ancestors, because largeforms have a short-term advantage but tend to be moreprone to extinction in the long run. The hypothesis of asmall ancestral size is also largely supported by the fossilrecord: most of the Cretaceous mammals are smaller thana few inches, whereas post-KT deposits include numerouslarge mammals (Luo 2007; Smith et al. 2010).

Paleontology, however, is not conclusive regarding the an-cestry of extant mammals due to the difficulty of linkingCretaceous and Paleocene fossils to modern lineages(Archibald et al. 2001; Asher et al. 2005). Genomic data pro-vide an attractive opportunity to characterize the ancestralfeatures of extant species: genome dynamics is imprinted byspecies life-history traits (Nikolaev et al. 2007; Nabholz et al.2008; Romiguier et al. 2010), and ancestral genome characterscan be reconstructed by phylogenetic methods (Galtier et al.1999; Blanchette et al. 2004; Boussau et al. 2009; Lartillot andPoujol 2011). If molecular evolution can be traced back to thelast common ancestor of extant placentals, then we couldpotentially learn about its macroscopic characteristics, eventhough this ancestor is not physically observable becausemissing from the fossil record.

In mammals, two genomic variables are known to correlatewith species life-history traits. First, species longevity and bodymass influence the ratio of nonsynonymous (= amino acidchanging, dN) to synonymous (dS) nucleotide substitutionrates. It has been shown that large and long-lived speciesdisplay a higher dN/dS ratio, on average, than small andshort-lived ones, presumably because of the smaller averagepopulation sizes, and hence the less effective purifying selec-tion, in long-lived animals (Nikolaev et al. 2007). Second, largespecies tend to show a lower GC3 (percentage of G and C atthe third position of codons) than small species. This effect issupposed to be caused by GC-biased gene conversion (Duretand Galtier 2009), a mechanism by which a biased DNA-repair process favors G and C alleles during meiotic recom-bination. Because short-lived species experience a higher rateof meiosis per time unit, their genome shows a faster diver-gence in gene GC3 and an increase in average GC3 (Romiguieret al. 2010).

Here, we quantify the influence of species longevity ongene coding sequence evolutionary dynamics (dN/dS ratioand GC3 divergence) in modern placental mammals. Thenwe reconstruct ancestral gene sequences using nonhomoge-neous phylogenetic models and estimate the dN/dS and GC3dynamics in the deepest branches of the mammalian tree. Onthe basis of the existing correlation between genomic pro-cesses and traits, we estimate the maximal life span of earlyplacental mammals. Our analysis suggests that the ancestorsof living placentals had a longevity and body mass similar tocurrent primates, cetartiodactyls, or carnivores but differed

Article

FastT

rack

! The Author 2012. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, pleasee-mail: journals.permissions@oup.com

Mol. Biol. Evol. 30(1):5–13 doi:10.1093/molbev/mss211 Advance Access publication September 4, 2012 5

at UniversitÃ

© &

EPFL Lausanne on January 21, 2013http://m

be.oxfordjournals.org/D

ownloaded from

An issue with sequence phylogenies

•Can be ambiguous if not enough information.

Ambiguities in mutationsancestral sequence species 1. species 2. What happened

Ambiguities in mutationsancestral sequence species 1. species 2. What happened

Ambiguities in mutationsancestral sequence species 1. species 2. What happened

Ambiguities in mutationsancestral sequence species 1. species 2. What happened

Ambiguities in mutationsancestral sequence species 1. species 2. What happened

Ambiguities in mutationsancestral sequence species 1. species 2. What happened

Ambiguities in mutationsancestral sequence species 1. species 2. What happened

•Can be ambiguous if not enough information.

An issue with sequence phylogenies

•Can be ambiguous if not enough information.

An issue with sequence phylogenies

•Used to be expensive.

•Can be ambiguous if not enough information.

An issue with sequence phylogenies

•Used to be expensive.

•Mitochondrial gene vs. nuclear gene. Several genes?

•Can be ambiguous if not enough information.

An issue with sequence phylogenies

•Whole genome sequencing is now dirt cheap! No longer a problem! (for establishing relationships in past 200-400 million years...)

•Used to be expensive.

•Mitochondrial gene vs. nuclear gene. Several genes?

3. DNA sequence clarifies current evolutionary contexts

Three-spined stickleback Gasterosteus aculeatus

Bill Cresko et al; David Kingsley et al

Different amounts of armor plating

Example 1.

Three-spined stickleback Gasterosteus aculeatus

in Saltwater:

in Freshwater:

Bill Cresko et al; David Kingsley et al

Different amounts of armor plating

Example 1.

Independent colonization events

less than 10,000 years ago

in Saltwater:

in Freshwater:

conservative (and unbiased) nature with which SNPs are calledusing our methodology (see Methods), and additional sequencingof these samples may increase the number of SNPs identified.Furthermore, in agreement with the hypothesis that freshwaterpopulations in this region have been derived post-glacially fromoceanic populations [49,55,65,79], global genetic diversity mea-sures are increased only slightly when combining pairs ofpopulations whether they are both oceanic, both freshwater, orone of each (Table 2).

Our data support the hypothesis that oceanic sticklebackpopulations have few barriers to dispersal, relatively large amountsof gene flow, and little population genetic subdivision[55,57,59,60,103,104]. Rabbit Slough and Resurrection Bay, thetwo oceanic populations in our study, are the most geographicallydistant from one another (.1000 km as the fish swims). Despitethis distance, the oceanic populations show the least amount ofdifferentiation between them (FST = 0.0076; Table 2). In contrast,higher values of FST were observed in pairwise comparisonsamong freshwater populations and between freshwater andoceanic populations (0.05–0.15), which is generally interpretedas low to moderate amounts of population structuring (Table 2).

The freshwater populations, despite their younger age, are moredivergent both from the oceanic ancestral populations and fromeach other, consistent with our supposition that they representindependent colonizations from the ancestral oceanic population.These results are remarkably similar to results obtained previouslyfrom some of these same populations using a small number ofmicrosatellite and mtDNA markers [55]. This combination oflarge amounts of genetic variation and overall low-to-moderatedifferentiation between populations, coupled with recent and rapidphenotypic evolution in the freshwater populations, presents anideal situation for identifying genomic regions that have respondedto various kinds of natural selection.

Patterns of genetic diversity distributed across thegenome

To assess genome-wide patterns we examined mean nucleotidediversity (p) and heterozygosity (H) using a Gaussian kernelsmoothing function across each linkage group (Figure 4 and FigureS1). Although the overall mean diversity and heterozygosity valuesare 0.00336 and 0.00187, respectively, values vary widely acrossthe genome. Nucleotide diversity within genomic regions rangesfrom 0.0003 to over 0.01, whereas heterozygosity values rangefrom 0.0001 to 0.0083. This variation in diversity across thegenome provides important clues to the evolutionary processesthat are maintaining genetic diversity. For example, whileexpected (p) and observed (H) heterozygosity largely correspond,they differ at a few genomic regions (e.g., on Linkage Group XI).Genomic regions that exhibit significantly (p,1025) low levels ofdiversity and heterozygosity (e.g. on LG II and V, Figure 4and Figure S1) may be the result of low mutation rate,low recombination rate, purifying or positive selection that isconsistent across populations, or some combination of factors[9,36,105–107].

In contrast, other genomic regions, such as those on LG III andXIII (Figure 4), show very high levels of both diversity andheterozygosity. The most striking such region, found near the endof LG III, corresponds precisely with a region of reduceddifferentiation among populations (Figure 5). This suggests thepresence of balancing selection maintaining a common pool ofgenetic variation at this genomic region within and amongpopulations. To further investigate the pattern of increased geneticvariation on LG III, we delineated a region from 14.8 to 16.1 Mb(Figure 5; see Methods). Within the corresponding 1.3-Mb interval inthe published stickleback genome are several candidate targets ofbalancing selection, namely genes implicated in the first line ofdefense against pathogens [108]: ZEB1 (ENSGACG00000017648),and two adjacent APOL genes (ENSGACG00000017778, EN-SGACG00000017779). Supporting the importance of this region inimmune response, there are also orthologs of several inflammationpathway genes: LTB4R (ENSGACG00000017812), SHARPIN(ENSGACG00000017834), and CEBPD (ENSGACG00000017927)[109–111]. The region of significantly elevated nucleotide diversityon LG XIII (18.1–19.1 Mb) also contains candidate targets ofbalancing selection including a TRIM14 (ENSGACG00000014283)and three TRIM35 genes (ENSGACG00000014250, ENSG-ACG00000014251, ENSGACG00000014403). Many members ofthis large gene family have been implicated in innate immuneresponse (reviewed in [112]), and one gene, TRIM5alpha, bears thesignature of balancing selection in primates [113]. The sticklebackTRIM cluster on LG XIII provides a second example of balancingselection acting at a TRIM locus.

Evidence for balancing selection on Major HistoCompatibility(MHC) loci is somewhat weaker. An MHC Class II gene(ENSGACG00000017967) falls nearly 580 kb outside the interval

Figure 1. Location of oceanic and freshwater populationsexamined. Threespine stickleback were sampled from three freshwa-ter (Bear Paw Lake [BP], Boot Lake [BL], Mud Lake [ML]) and two oceanic(Rabbit Slough [RS], Resurrection Bay [RB]) populations in south centralAlaska, USA (see inset). The three freshwater populations occur indifferent drainages and are separated by barriers to dispersal, andprevious evidence supports the hypothesis that they representindependent colonization events from ancestral oceanic populations[49].doi:10.1371/journal.pgen.1000862.g001

Population Genomics in Stickleback

PLoS Genetics | www.plosgenetics.org 3 February 2010 | Volume 6 | Issue 2 | e1000862

F

S

F

F

S

S = Saltwater

F = Freshwater

Bill Cresko et al;

RAD = Restriction-site Associated DNA sequencing

each locus sequenced 5–10 times per fish.

Bill Cresko et al;

conservative (and unbiased) nature with which SNPs are calledusing our methodology (see Methods), and additional sequencingof these samples may increase the number of SNPs identified.Furthermore, in agreement with the hypothesis that freshwaterpopulations in this region have been derived post-glacially fromoceanic populations [49,55,65,79], global genetic diversity mea-sures are increased only slightly when combining pairs ofpopulations whether they are both oceanic, both freshwater, orone of each (Table 2).

Our data support the hypothesis that oceanic sticklebackpopulations have few barriers to dispersal, relatively large amountsof gene flow, and little population genetic subdivision[55,57,59,60,103,104]. Rabbit Slough and Resurrection Bay, thetwo oceanic populations in our study, are the most geographicallydistant from one another (.1000 km as the fish swims). Despitethis distance, the oceanic populations show the least amount ofdifferentiation between them (FST = 0.0076; Table 2). In contrast,higher values of FST were observed in pairwise comparisonsamong freshwater populations and between freshwater andoceanic populations (0.05–0.15), which is generally interpretedas low to moderate amounts of population structuring (Table 2).

The freshwater populations, despite their younger age, are moredivergent both from the oceanic ancestral populations and fromeach other, consistent with our supposition that they representindependent colonizations from the ancestral oceanic population.These results are remarkably similar to results obtained previouslyfrom some of these same populations using a small number ofmicrosatellite and mtDNA markers [55]. This combination oflarge amounts of genetic variation and overall low-to-moderatedifferentiation between populations, coupled with recent and rapidphenotypic evolution in the freshwater populations, presents anideal situation for identifying genomic regions that have respondedto various kinds of natural selection.

Patterns of genetic diversity distributed across thegenome

To assess genome-wide patterns we examined mean nucleotidediversity (p) and heterozygosity (H) using a Gaussian kernelsmoothing function across each linkage group (Figure 4 and FigureS1). Although the overall mean diversity and heterozygosity valuesare 0.00336 and 0.00187, respectively, values vary widely acrossthe genome. Nucleotide diversity within genomic regions rangesfrom 0.0003 to over 0.01, whereas heterozygosity values rangefrom 0.0001 to 0.0083. This variation in diversity across thegenome provides important clues to the evolutionary processesthat are maintaining genetic diversity. For example, whileexpected (p) and observed (H) heterozygosity largely correspond,they differ at a few genomic regions (e.g., on Linkage Group XI).Genomic regions that exhibit significantly (p,1025) low levels ofdiversity and heterozygosity (e.g. on LG II and V, Figure 4and Figure S1) may be the result of low mutation rate,low recombination rate, purifying or positive selection that isconsistent across populations, or some combination of factors[9,36,105–107].

In contrast, other genomic regions, such as those on LG III andXIII (Figure 4), show very high levels of both diversity andheterozygosity. The most striking such region, found near the endof LG III, corresponds precisely with a region of reduceddifferentiation among populations (Figure 5). This suggests thepresence of balancing selection maintaining a common pool ofgenetic variation at this genomic region within and amongpopulations. To further investigate the pattern of increased geneticvariation on LG III, we delineated a region from 14.8 to 16.1 Mb(Figure 5; see Methods). Within the corresponding 1.3-Mb interval inthe published stickleback genome are several candidate targets ofbalancing selection, namely genes implicated in the first line ofdefense against pathogens [108]: ZEB1 (ENSGACG00000017648),and two adjacent APOL genes (ENSGACG00000017778, EN-SGACG00000017779). Supporting the importance of this region inimmune response, there are also orthologs of several inflammationpathway genes: LTB4R (ENSGACG00000017812), SHARPIN(ENSGACG00000017834), and CEBPD (ENSGACG00000017927)[109–111]. The region of significantly elevated nucleotide diversityon LG XIII (18.1–19.1 Mb) also contains candidate targets ofbalancing selection including a TRIM14 (ENSGACG00000014283)and three TRIM35 genes (ENSGACG00000014250, ENSG-ACG00000014251, ENSGACG00000014403). Many members ofthis large gene family have been implicated in innate immuneresponse (reviewed in [112]), and one gene, TRIM5alpha, bears thesignature of balancing selection in primates [113]. The sticklebackTRIM cluster on LG XIII provides a second example of balancingselection acting at a TRIM locus.

Evidence for balancing selection on Major HistoCompatibility(MHC) loci is somewhat weaker. An MHC Class II gene(ENSGACG00000017967) falls nearly 580 kb outside the interval

Figure 1. Location of oceanic and freshwater populationsexamined. Threespine stickleback were sampled from three freshwa-ter (Bear Paw Lake [BP], Boot Lake [BL], Mud Lake [ML]) and two oceanic(Rabbit Slough [RS], Resurrection Bay [RB]) populations in south centralAlaska, USA (see inset). The three freshwater populations occur indifferent drainages and are separated by barriers to dispersal, andprevious evidence supports the hypothesis that they representindependent colonization events from ancestral oceanic populations[49].doi:10.1371/journal.pgen.1000862.g001

Population Genomics in Stickleback

PLoS Genetics | www.plosgenetics.org 3 February 2010 | Volume 6 | Issue 2 | e1000862

F

S

F

F

S

S = Saltwater

F = Freshwater

20 fish per population

45,789 SNPs identified

Differentiation between populations (FST)

Saltwater vs.

Saltwater

Figure 6. Genome-wide differentiation among populations. FST across the genome, with colored bars indicating significantly elevated(p#1025, blue; p#1027, red) and reduced (p#1025, green) values. Vertical gray shading indicates boundaries of the linkage groups and unassembledscaffolds, and gold shading indicates the nine peaks of substantial population differentiation discussed in the text. (A) FST between the two oceanicpopulations (RS and RB; note that no regions of FST are significantly elevated or reduced). (B,C,D) Differentiation of each single freshwater populationfrom the two oceanic populations, shown as the mean of the two pairwise comparisons (with RS and RB): (B) BP, (C) BL, (D) ML. Colored bars in eachplot represent regions where both pairwise comparisons exceeded the corresponding significance threshold. (E) Overall population differentiationbetween the oceanic and freshwater populations. (F) Differentiation among the three freshwater populations (BP, BL, ML).doi:10.1371/journal.pgen.1000862.g006

Population Genomics in Stickleback

PLoS Genetics | www.plosgenetics.org 8 February 2010 | Volume 6 | Issue 2 | e1000862

FST bewteen 2 populations: 0 = populations have same alleles in similar frequencies 1 = populations have completely different alleles

Bill Cresko et al; David Kingsley et al

Differentiation between populations (FST)

Saltwater vs.

Saltwater

Figure 6. Genome-wide differentiation among populations. FST across the genome, with colored bars indicating significantly elevated(p#1025, blue; p#1027, red) and reduced (p#1025, green) values. Vertical gray shading indicates boundaries of the linkage groups and unassembledscaffolds, and gold shading indicates the nine peaks of substantial population differentiation discussed in the text. (A) FST between the two oceanicpopulations (RS and RB; note that no regions of FST are significantly elevated or reduced). (B,C,D) Differentiation of each single freshwater populationfrom the two oceanic populations, shown as the mean of the two pairwise comparisons (with RS and RB): (B) BP, (C) BL, (D) ML. Colored bars in eachplot represent regions where both pairwise comparisons exceeded the corresponding significance threshold. (E) Overall population differentiationbetween the oceanic and freshwater populations. (F) Differentiation among the three freshwater populations (BP, BL, ML).doi:10.1371/journal.pgen.1000862.g006

Population Genomics in Stickleback

PLoS Genetics | www.plosgenetics.org 8 February 2010 | Volume 6 | Issue 2 | e1000862

Freshwater vs.

Freshwater

Figure 6. Genome-wide differentiation among populations. FST across the genome, with colored bars indicating significantly elevated(p#1025, blue; p#1027, red) and reduced (p#1025, green) values. Vertical gray shading indicates boundaries of the linkage groups and unassembledscaffolds, and gold shading indicates the nine peaks of substantial population differentiation discussed in the text. (A) FST between the two oceanicpopulations (RS and RB; note that no regions of FST are significantly elevated or reduced). (B,C,D) Differentiation of each single freshwater populationfrom the two oceanic populations, shown as the mean of the two pairwise comparisons (with RS and RB): (B) BP, (C) BL, (D) ML. Colored bars in eachplot represent regions where both pairwise comparisons exceeded the corresponding significance threshold. (E) Overall population differentiationbetween the oceanic and freshwater populations. (F) Differentiation among the three freshwater populations (BP, BL, ML).doi:10.1371/journal.pgen.1000862.g006

Population Genomics in Stickleback

PLoS Genetics | www.plosgenetics.org 8 February 2010 | Volume 6 | Issue 2 | e1000862FST bewteen 2 populations: 0 = populations have same alleles in similar frequencies

1 = populations have completely different allelesBill Cresko et al; David Kingsley et al

Differentiation between populations (FST)

Saltwater vs.

Saltwater

Figure 6. Genome-wide differentiation among populations. FST across the genome, with colored bars indicating significantly elevated(p#1025, blue; p#1027, red) and reduced (p#1025, green) values. Vertical gray shading indicates boundaries of the linkage groups and unassembledscaffolds, and gold shading indicates the nine peaks of substantial population differentiation discussed in the text. (A) FST between the two oceanicpopulations (RS and RB; note that no regions of FST are significantly elevated or reduced). (B,C,D) Differentiation of each single freshwater populationfrom the two oceanic populations, shown as the mean of the two pairwise comparisons (with RS and RB): (B) BP, (C) BL, (D) ML. Colored bars in eachplot represent regions where both pairwise comparisons exceeded the corresponding significance threshold. (E) Overall population differentiationbetween the oceanic and freshwater populations. (F) Differentiation among the three freshwater populations (BP, BL, ML).doi:10.1371/journal.pgen.1000862.g006

Population Genomics in Stickleback

PLoS Genetics | www.plosgenetics.org 8 February 2010 | Volume 6 | Issue 2 | e1000862

Freshwater vs.

Freshwater

Figure 6. Genome-wide differentiation among populations. FST across the genome, with colored bars indicating significantly elevated(p#1025, blue; p#1027, red) and reduced (p#1025, green) values. Vertical gray shading indicates boundaries of the linkage groups and unassembledscaffolds, and gold shading indicates the nine peaks of substantial population differentiation discussed in the text. (A) FST between the two oceanicpopulations (RS and RB; note that no regions of FST are significantly elevated or reduced). (B,C,D) Differentiation of each single freshwater populationfrom the two oceanic populations, shown as the mean of the two pairwise comparisons (with RS and RB): (B) BP, (C) BL, (D) ML. Colored bars in eachplot represent regions where both pairwise comparisons exceeded the corresponding significance threshold. (E) Overall population differentiationbetween the oceanic and freshwater populations. (F) Differentiation among the three freshwater populations (BP, BL, ML).doi:10.1371/journal.pgen.1000862.g006

Population Genomics in Stickleback

PLoS Genetics | www.plosgenetics.org 8 February 2010 | Volume 6 | Issue 2 | e1000862

Freshwater vs.

Saltwater

Figure 6. Genome-wide differentiation among populations. FST across the genome, with colored bars indicating significantly elevated(p#1025, blue; p#1027, red) and reduced (p#1025, green) values. Vertical gray shading indicates boundaries of the linkage groups and unassembledscaffolds, and gold shading indicates the nine peaks of substantial population differentiation discussed in the text. (A) FST between the two oceanicpopulations (RS and RB; note that no regions of FST are significantly elevated or reduced). (B,C,D) Differentiation of each single freshwater populationfrom the two oceanic populations, shown as the mean of the two pairwise comparisons (with RS and RB): (B) BP, (C) BL, (D) ML. Colored bars in eachplot represent regions where both pairwise comparisons exceeded the corresponding significance threshold. (E) Overall population differentiationbetween the oceanic and freshwater populations. (F) Differentiation among the three freshwater populations (BP, BL, ML).doi:10.1371/journal.pgen.1000862.g006

Population Genomics in Stickleback

PLoS Genetics | www.plosgenetics.org 8 February 2010 | Volume 6 | Issue 2 | e1000862

FST bewteen 2 populations: 0 = populations have same alleles in similar frequencies 1 = populations have completely different alleles

Bill Cresko et al; David Kingsley et al

Nine identified regions• Include 590 genes:

• 31 of these known to or likely to affect morphology or osmoregulation

• some previously identified via crosses; most new• finer-scale follow-up required?

Nine identified regions• Include 590 genes:

• 31 of these known to or likely to affect morphology or osmoregulation

• some previously identified via crosses; most new• finer-scale follow-up required?

• Different regions, different mode of evolution:

Nine identified regions• Include 590 genes:

• 31 of these known to or likely to affect morphology or osmoregulation

• some previously identified via crosses; most new• finer-scale follow-up required?

• Different regions, different mode of evolution:• hard sweep: the same allele is selected to fixation in all

freshwater populations

Nine identified regions• Include 590 genes:

• 31 of these known to or likely to affect morphology or osmoregulation

• some previously identified via crosses; most new• finer-scale follow-up required?

• Different regions, different mode of evolution:• hard sweep: the same allele is selected to fixation in all

freshwater populations• soft sweep: a different allele is selected to fixation in each

freshwater population

Nine identified regions• Include 590 genes:

• 31 of these known to or likely to affect morphology or osmoregulation

• some previously identified via crosses; most new• finer-scale follow-up required?

• Different regions, different mode of evolution:• hard sweep: the same allele is selected to fixation in all

freshwater populations• soft sweep: a different allele is selected to fixation in each

freshwater population• balancing selection: selection for heterozygosity

Little fire ant Wasmannia

Little fire ant Wasmannia

reproduction (that is, by ameiotic parthenogenesis). In 33 of the 34nests, all queens (n ! 135) and gynes (n ! 9) cohabiting in the samenest shared an identical genotype at each of the 11 loci (Table 1 andFig. 1). The single exception was nest B-12, in which queens differedat 1 of the 11 loci: four queens were heterozygous at Waur-2164and the remaining three queens were homozygous for one of thetwo alleles. This variation probably reflects a mutation or recombi-nation event in one queen followed by clonal reproduction withinthe nest. The history of this genetic change could be reconstructedfrom the genotypes of queens collected in neighbouring nests (Figs 1and 2). Nine queens from two neighbouring nests (B-11 and B-13)had the same genotype as the four heterozygous queens for locusWaur-2164, indicating that the mutation or recombination eventprobably was from a heterozygote to a homozygote queen. The threehomozygote queens from nest B-12 had a unique genotype in thepopulation, which further supports this interpretation.A comparison between nests supports the view of restricted female

gene flow, with budding being the main mode of colony formation.Within three of the five sites of collection (A, C and D) all queens hadthe same genotype at the 11 loci (Fig. 2). In one of the two other sites(B), all queens from 8 of the 17 nests also had an identical genotype,whereas in the other site (E) the queen genotypes were different in thethree nests sampled. Taken together, these data indicate that queensbelonging to the same lineage of clonally produced individualsfrequently head closely located nests. Moreover, genetic differen-tiation between sites was very strong, with a single occurrence ofgenotypes shared between sites (the eight queens of nest E-3 hadgenotypes identical to the most common genotype found at site B),showing that gene flow by females is extremely restricted.In stark contrast to reproductive females, the genotypic analyses

revealed that workers are produced by normal sexual reproduction(Table 1). Over all 31 queenright nests, each of the 248 genotypedworkers had, at seven or more loci, one allele that was absent inqueens of their nest. Moreover, the 232 workers from the 29 nests inwhich the sperm in the queen’s spermathecae was successfullyobtained had all genotypes consistent with those expected undersexual reproduction between the two parental genomes.The genetic analyses of the sperm collected in the queens’

Table 1 | Genotypes of queens (Q), their mates (M) and workers (w) in one nest (E-3) at each of the 11 microsatellite loci

Individual Waur-225 Waur-275 Waur-418 Waur-566 Waur-680 Waur-716 Waur-730 Waur-1166 Waur-2164 Waur-3176 Waur-1gam

QueensQ-1 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-2 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-3 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-4 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-5 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-6 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-7 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-8 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298MalesM-1 269 107 118 265 187 192 214 95 320 244 282M-2 269 107 118 265 187 192 214 95 320 244 282M-3 269 107 118 265 187 192 214 95 320 244 282M-4 269 107 118 265 187 192 214 95 320 244 282M-5 269 107 118 265 187 192 214 95 320 244 282M-6 269 107 118 265 187 192 214 95 320 244 282M-7 269 107 118 265 187 192 214 95 320 244 282M-8 269 107 118 265 187 192 214 95 320 244 282Workersw-1 223 269 115 107 112 118 263 265 171 187 198 192 160 214 95 95 306 320 230 244 298 282w-2 225 269 115 107 100 118 263 265 171 187 184 192 158 214 95 95 298 320 230 244 288 282w-3 223 269 105 107 112 118 263 265 171 187 198 192 160 214 97 95 298 320 230 244 298 282w-4 225 269 115 107 100 118 263 265 171 187 184 192 158 214 97 95 306 320 230 244 288 282w-5 223 269 105 107 100 118 263 265 171 187 198 192 158 214 97 95 306 320 230 244 298 282w-6 225 269 115 107 112 118 263 265 171 187 184 192 160 214 97 95 306 320 230 244 288 282w-7 223 269 105 107 100 118 263 265 171 187 184 192 158 214 97 95 306 320 230 244 298 282w-8 225 269 115 107 112 118 263 265 171 187 184 192 158 214 97 95 298 320 230 244 288 282

The identities of mates were determined by the sperm collected in the queen’s spermathecae. Queens and males’ genotypes illustrate their clonal production, whereas workers’ genotypes areconsistent with normal sexual reproduction. Paternal alleles are in italics.

Figure 2 | Neighbour-joining dendrogram of the genetic (allele-shared)distances between queens (Q), gynes (G) and male sperms (M) collectedover all the five sites (A–E). The collection number of each nest is givenwith the letter of the site (see Fig. 1 for details). The number of individualssharing the same genotype (n) is given for all nests.

NATURE|Vol 435|30 June 2005 LETTERS

1231© 2005 Nature Publishing Group

reproduction (that is, by ameiotic parthenogenesis). In 33 of the 34nests, all queens (n ! 135) and gynes (n ! 9) cohabiting in the samenest shared an identical genotype at each of the 11 loci (Table 1 andFig. 1). The single exception was nest B-12, in which queens differedat 1 of the 11 loci: four queens were heterozygous at Waur-2164and the remaining three queens were homozygous for one of thetwo alleles. This variation probably reflects a mutation or recombi-nation event in one queen followed by clonal reproduction withinthe nest. The history of this genetic change could be reconstructedfrom the genotypes of queens collected in neighbouring nests (Figs 1and 2). Nine queens from two neighbouring nests (B-11 and B-13)had the same genotype as the four heterozygous queens for locusWaur-2164, indicating that the mutation or recombination eventprobably was from a heterozygote to a homozygote queen. The threehomozygote queens from nest B-12 had a unique genotype in thepopulation, which further supports this interpretation.A comparison between nests supports the view of restricted female

gene flow, with budding being the main mode of colony formation.Within three of the five sites of collection (A, C and D) all queens hadthe same genotype at the 11 loci (Fig. 2). In one of the two other sites(B), all queens from 8 of the 17 nests also had an identical genotype,whereas in the other site (E) the queen genotypes were different in thethree nests sampled. Taken together, these data indicate that queensbelonging to the same lineage of clonally produced individualsfrequently head closely located nests. Moreover, genetic differen-tiation between sites was very strong, with a single occurrence ofgenotypes shared between sites (the eight queens of nest E-3 hadgenotypes identical to the most common genotype found at site B),showing that gene flow by females is extremely restricted.In stark contrast to reproductive females, the genotypic analyses

revealed that workers are produced by normal sexual reproduction(Table 1). Over all 31 queenright nests, each of the 248 genotypedworkers had, at seven or more loci, one allele that was absent inqueens of their nest. Moreover, the 232 workers from the 29 nests inwhich the sperm in the queen’s spermathecae was successfullyobtained had all genotypes consistent with those expected undersexual reproduction between the two parental genomes.The genetic analyses of the sperm collected in the queens’

Table 1 | Genotypes of queens (Q), their mates (M) and workers (w) in one nest (E-3) at each of the 11 microsatellite loci

Individual Waur-225 Waur-275 Waur-418 Waur-566 Waur-680 Waur-716 Waur-730 Waur-1166 Waur-2164 Waur-3176 Waur-1gam

QueensQ-1 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-2 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-3 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-4 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-5 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-6 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-7 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-8 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298MalesM-1 269 107 118 265 187 192 214 95 320 244 282M-2 269 107 118 265 187 192 214 95 320 244 282M-3 269 107 118 265 187 192 214 95 320 244 282M-4 269 107 118 265 187 192 214 95 320 244 282M-5 269 107 118 265 187 192 214 95 320 244 282M-6 269 107 118 265 187 192 214 95 320 244 282M-7 269 107 118 265 187 192 214 95 320 244 282M-8 269 107 118 265 187 192 214 95 320 244 282Workersw-1 223 269 115 107 112 118 263 265 171 187 198 192 160 214 95 95 306 320 230 244 298 282w-2 225 269 115 107 100 118 263 265 171 187 184 192 158 214 95 95 298 320 230 244 288 282w-3 223 269 105 107 112 118 263 265 171 187 198 192 160 214 97 95 298 320 230 244 298 282w-4 225 269 115 107 100 118 263 265 171 187 184 192 158 214 97 95 306 320 230 244 288 282w-5 223 269 105 107 100 118 263 265 171 187 198 192 158 214 97 95 306 320 230 244 298 282w-6 225 269 115 107 112 118 263 265 171 187 184 192 160 214 97 95 306 320 230 244 288 282w-7 223 269 105 107 100 118 263 265 171 187 184 192 158 214 97 95 306 320 230 244 298 282w-8 225 269 115 107 112 118 263 265 171 187 184 192 158 214 97 95 298 320 230 244 288 282

The identities of mates were determined by the sperm collected in the queen’s spermathecae. Queens and males’ genotypes illustrate their clonal production, whereas workers’ genotypes areconsistent with normal sexual reproduction. Paternal alleles are in italics.

Figure 2 | Neighbour-joining dendrogram of the genetic (allele-shared)distances between queens (Q), gynes (G) and male sperms (M) collectedover all the five sites (A–E). The collection number of each nest is givenwith the letter of the site (see Fig. 1 for details). The number of individualssharing the same genotype (n) is given for all nests.

NATURE|Vol 435|30 June 2005 LETTERS

1231© 2005 Nature Publishing Group

reproduction (that is, by ameiotic parthenogenesis). In 33 of the 34nests, all queens (n ! 135) and gynes (n ! 9) cohabiting in the samenest shared an identical genotype at each of the 11 loci (Table 1 andFig. 1). The single exception was nest B-12, in which queens differedat 1 of the 11 loci: four queens were heterozygous at Waur-2164and the remaining three queens were homozygous for one of thetwo alleles. This variation probably reflects a mutation or recombi-nation event in one queen followed by clonal reproduction withinthe nest. The history of this genetic change could be reconstructedfrom the genotypes of queens collected in neighbouring nests (Figs 1and 2). Nine queens from two neighbouring nests (B-11 and B-13)had the same genotype as the four heterozygous queens for locusWaur-2164, indicating that the mutation or recombination eventprobably was from a heterozygote to a homozygote queen. The threehomozygote queens from nest B-12 had a unique genotype in thepopulation, which further supports this interpretation.A comparison between nests supports the view of restricted female

gene flow, with budding being the main mode of colony formation.Within three of the five sites of collection (A, C and D) all queens hadthe same genotype at the 11 loci (Fig. 2). In one of the two other sites(B), all queens from 8 of the 17 nests also had an identical genotype,whereas in the other site (E) the queen genotypes were different in thethree nests sampled. Taken together, these data indicate that queensbelonging to the same lineage of clonally produced individualsfrequently head closely located nests. Moreover, genetic differen-tiation between sites was very strong, with a single occurrence ofgenotypes shared between sites (the eight queens of nest E-3 hadgenotypes identical to the most common genotype found at site B),showing that gene flow by females is extremely restricted.In stark contrast to reproductive females, the genotypic analyses

revealed that workers are produced by normal sexual reproduction(Table 1). Over all 31 queenright nests, each of the 248 genotypedworkers had, at seven or more loci, one allele that was absent inqueens of their nest. Moreover, the 232 workers from the 29 nests inwhich the sperm in the queen’s spermathecae was successfullyobtained had all genotypes consistent with those expected undersexual reproduction between the two parental genomes.The genetic analyses of the sperm collected in the queens’

Table 1 | Genotypes of queens (Q), their mates (M) and workers (w) in one nest (E-3) at each of the 11 microsatellite loci

Individual Waur-225 Waur-275 Waur-418 Waur-566 Waur-680 Waur-716 Waur-730 Waur-1166 Waur-2164 Waur-3176 Waur-1gam

QueensQ-1 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-2 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-3 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-4 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-5 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-6 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-7 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298Q-8 223 225 105 115 100 112 263 263 171 171 184 198 158 160 95 97 298 306 230 230 288 298MalesM-1 269 107 118 265 187 192 214 95 320 244 282M-2 269 107 118 265 187 192 214 95 320 244 282M-3 269 107 118 265 187 192 214 95 320 244 282M-4 269 107 118 265 187 192 214 95 320 244 282M-5 269 107 118 265 187 192 214 95 320 244 282M-6 269 107 118 265 187 192 214 95 320 244 282M-7 269 107 118 265 187 192 214 95 320 244 282M-8 269 107 118 265 187 192 214 95 320 244 282Workersw-1 223 269 115 107 112 118 263 265 171 187 198 192 160 214 95 95 306 320 230 244 298 282w-2 225 269 115 107 100 118 263 265 171 187 184 192 158 214 95 95 298 320 230 244 288 282w-3 223 269 105 107 112 118 263 265 171 187 198 192 160 214 97 95 298 320 230 244 298 282w-4 225 269 115 107 100 118 263 265 171 187 184 192 158 214 97 95 306 320 230 244 288 282w-5 223 269 105 107 100 118 263 265 171 187 198 192 158 214 97 95 306 320 230 244 298 282w-6 225 269 115 107 112 118 263 265 171 187 184 192 160 214 97 95 306 320 230 244 288 282w-7 223 269 105 107 100 118 263 265 171 187 184 192 158 214 97 95 306 320 230 244 298 282w-8 225 269 115 107 112 118 263 265 171 187 184 192 158 214 97 95 298 320 230 244 288 282

The identities of mates were determined by the sperm collected in the queen’s spermathecae. Queens and males’ genotypes illustrate their clonal production, whereas workers’ genotypes areconsistent with normal sexual reproduction. Paternal alleles are in italics.

Figure 2 | Neighbour-joining dendrogram of the genetic (allele-shared)distances between queens (Q), gynes (G) and male sperms (M) collectedover all the five sites (A–E). The collection number of each nest is givenwith the letter of the site (see Fig. 1 for details). The number of individualssharing the same genotype (n) is given for all nests.

NATURE|Vol 435|30 June 2005 LETTERS

1231© 2005 Nature Publishing Group

mate with their brothers inside the nest, and yet maintainheterozygosity in the queen and worker castes over anunlimited number of generations.

Surprisingly, the heterozygosity level of new queens iscompletely independent of the genealogical link betweenthe mother queen and her mate in this species, as thereis no mixing of the paternal and maternal lineages. Bycontrast, the level of worker heterozygosity depends onthe genetic similarity between the maternal and paternallineages, because workers are sexually produced. Butsince all queens and all males in our study populationhad the same genotype, queens were no more geneticallysimilar to their brothers than to any other male in thepopulation, and the level of worker heterozygosity wastherefore not affected by whether queens mated withtheir brothers or other males.

The particular reproductive system of P. longicornisprobably provides an important pre-adaptation allowingthis species to colonize new habitats. Paratrechina longicornisis arguably the most broadly dispersed of all ant species,distributed widely across the Old World and New Worldin both the Northern Hemisphere and Southern Hemi-sphere [25]. The multiple introductions of this specieshave probably been accompanied by repeated reductionsin population size during the founding stages of new popu-lations. However, the complete separation of the maternaland paternal genomes allows such founding events tooccur without any cost associated with inbreeding (i.e.mating between related individuals) in workers. Whateverthe population size, a high level of heterozygosity can bemaintained in workers as a result of the very stark geneticdifferences between the paternal and maternal lineages.Population bottlenecks only affect the allelic diversity atthe population level, without decreasing observed hetero-zygosity of queens and workers. In the studied population,there was actually no genotypic diversity within either thematernal or paternal clonal lineages, yet the observed het-erozygosity in workers was extremely high (across the 12microsatellite loci, observed heterozygosity was close to 1;electronic supplementary material, table S1).

Recent studies have reported similar reproductivestrategies involving clonal reproduction of both sexes in

two other ant species: Wasmannia auropunctata and Vollen-hovia emeryi [21,22]. On the basis of the results of ourstudy, it is likely that the mating system of these speciesalso translates into reduced effects of bottlenecks andsib mating on the level of heterozygosity of females. Inter-estingly, W. auropunctata is also an invasive ant, and recentstudies have shown that introduced populations oftenderive from a single queen introduction [30] and arecharacterized by the presence of a single queen and asingle male genotype. The presence of a single queenand a single male genotype in our P. longicornis populationis also compatible with the population being initiated bya single mated queen.

Interestingly, our study shows that P. longicornis queenslay male eggs that inherit no maternal genes. There are atleast two potential mechanisms that could lead to a malebeing clonally produced by his father. First, the maternalgenome could be eliminated after the egg was fertilized[21]. Indeed, other examples of selective elimination ofone parental genome have been reported in species suchas the parasitoid wasp Nasonia vitripennis [31,32] andthe waterfrog hybridogenetic species complex Rana escu-lenta [33,34], and as probable in some ants of the genusFormica [35,36]. In these cases, however, it is always thepaternal genome that is eliminated and not the maternalone. The alternative possibility is that females sometimeslay anucleated eggs that develop into haploid males whenfertilized [37]. Unfortunately, it is not possible to knowwhich eggs will give rise to males and females, and it willthus be very difficult to discriminate between these twopotential mechanisms.

In conclusion, this study demonstrates that P. longicor-nis uses an unusual mode of reproduction acting as apre-adaptation for situations where inbreeding and bottle-necking occur. Even if populations are initiated by a singlefemale, the clonal reproduction of both males and queensensures a high level of observed heterozyosity of workerswhen the founding queen and her mate have differentalleles. Moreover, this unusual system of reproductionmaintains similar levels of observed heterozygosity forqueens and for workers even if matings occur between sib-lings over several generations. This may be an importantfactor explaining why this species has been so successfulin invading new habitats.

We thank Sasithorn Hasin and Mingkwan Nipitwattanaphonfor collecting material in Thailand, Vicky Menetrey for hercontribution to lab work, and three anonymous reviewersfor their useful comments. This work was supported by anEMBO fellowship (M.P.), United States National ScienceFoundation grant DEB no. 0640690 (M.A.D.G.) and theSwiss National Science Foundation (L.K.).

REFERENCES1 Charlesworth, D. & Willlis, J. 2009 The genetics of

inbreeding depression. Nat. Rev. Genet. 10, 783–796.(doi:10.1038/nrg2664)

2 Whiting, P. W. 1939 Sex determination and reproductiveeconomy in Habrobracon. Genetics 24, 110–111.

3 Whiting, P. W. 1943 Multiple alleles in complementarysex determination of Habrobracon. Genetics 28, 365–382.

4 Butcher, R. D. J., Whitfield, W. G. F. & Hubbard, S. F. 2000Single-locus complementary sex determination in Diadegmachrysostictos (Gmelin) (Hymenoptera: Ichneumonidae).J. Hered. 91, 104–111. (doi:10.1093/jhered/91.2.104)

gynes workers males

queen mate

Figure 2. Clonal reproduction in queens and males. Thefigure summarizes the reproduction system of P. longicornisin the study population. Maternal (light) and paternal(dark) chromosomes are displayed. Contribution to thegenome of the offspring is indicated by arrows (dashedarrow represents the mother laying haploid eggs with noactual contribution to the genome).

2680 M. Pearcy et al. Sib mating without inbreeding

Proc. R. Soc. B (2011)

on January 21, 2013rspb.royalsocietypublishing.orgDownloaded from

Conservation

Summary

Summary

•1. DNA sequences change

Summary

•1. DNA sequences change•2. Evolutionary relationships

Summary

•1. DNA sequences change•2. Evolutionary relationships•3. Current evolutionary contexts