data and mathematical approaches to the neolithic transition

30
1 Data and mathematical approaches to the neolithic transition Joaquim Fort Universitat de Girona Catalonia, Spain FEPRE European project 3 rd annual workshop Girona, 16-18 March 2009

Upload: lynch

Post on 16-Jan-2016

22 views

Category:

Documents


0 download

DESCRIPTION

FEPRE European project 3 rd annual workshop Girona, 16-18 March 2009. Data and mathematical approaches to the neolithic transition. Joaquim Fort Universitat de Girona Catalonia, Spain. Plan of this talk. 1.Archaeological data: Speed versus neolithisation time - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Data and mathematical approaches  to the neolithic transition

1

Data and mathematical approaches to the neolithic transition

Joaquim FortUniversitat de Girona

Catalonia, Spain

FEPRE European project3rd annual workshopGirona, 16-18 March 2009

Page 2: Data and mathematical approaches  to the neolithic transition

2

1. Archaeological data: Speed versus neolithisation time

2. Mathematical models: Interpretation of the data

3. Archaeological data: Speed versus latitude

Plan of this talkPlan of this talk

Page 3: Data and mathematical approaches  to the neolithic transition

3

MotivationMotivation

If some hunter-gatherers become farmers:If some hunter-gatherers become farmers:· The front speed should be faster, and· The front speed should be faster, and· The · The neolithisation time*neolithisation time* should be shorter should be shorter

*Time necessary for the population of *Time necessary for the population of farmers to reach saturation densityfarmers to reach saturation density

· Theory: New J Phys (2008)· Comparison to archaeological data: this talk

Page 4: Data and mathematical approaches  to the neolithic transition

4

Neolithisation timeNeolithisation time

2400 yrData from Gkiasta et al., Antiquity (2003)Data from Gkiasta et al., Antiquity (2003) 2000 yr

Time necessary for the population Time necessary for the population of farmers to reach saturation densityof farmers to reach saturation density

Page 5: Data and mathematical approaches  to the neolithic transition

5

Legend! 2008_10_23a_fepre_context_ppnb9000CALyrBP Events

lakes

country

2008_10_23a

<VALUE>

<6.669 cal yr BP (>73 gen)

6.669 - 7.264 cal yr BP (73-54 gen)

7.265 - 7.859 cal yr BP (54-36 gen)

7.860 - 8.454 cal yr BP (36-17 gen)

>8.455 cal yr BP (<17 gen)

DATA INTERPOLATION: 903 FEPRE sites and 16 PPNB CONTEXT sites, all set to 9000 CAL yr BP2008_10_23a_fepre_context_ppnb9000CALyrBP.mxd2008_10_23b_fepre_context_ppnb9000CALyrBP.pdf

0 generations corresponds to 9000 cal yr BP

!

!

! ! !!!!!

! !!

! !! !!! !!! !!!! !! ! ! !!

! !! !!! !!! !! !

!!!

!! !! !!! !!! !!! !! !!! !!!!! ! !! ! ! !! !!! ! ! !! ! !! !! !!!!! !! ! !!!! !! !! !! !

!!! !! ! !!! !! !! !!! !!!! ! !! ! ! !!! ! ! !! !! !! ! !! !!! ! !!!! !!!! !! ! !!

! !! !!! !!! !! !! !! !!! ! !!!! !! ! ! !! !! ! ! !! ! !!! !! ! !! !!! ! ! !!! !! ! !!!! ! !! !!! ! !! !! ! !! !!!!!! ! ! !! ! !! !!! ! !! !! ! ! !! !!! ! !! !!! !! ! !!! !! !! !!!! ! ! ! ! ! !! !!! ! ! !! !!! ! !! ! !!!!! !! !! ! ! !!! !! ! !!! !! !!! ! !!!!! ! !!! ! !! ! !!! !! !! !! !!! !! ! !!! !! ! !! !! ! !! ! !!! ! ! !! !! ! ! !! !! !!!! ! !! ! !!! !!! ! !!! ! !! !! !! ! !! !! !!!

! !!! ! ! !! !! !! ! !!!! !! ! !! !!! ! !! ! !! ! ! !!! !! ! !!! ! !! !!!! ! !!! !! ! !! !! ! !!! !! !! ! ! !! !! ! !!! ! !!! !! ! ! ! !! !! !!! ! !! ! !!!! ! !!!! !! !! ! !! ! !!! ! !! ! ! !!! !! !!! !!!!!!!!!!! !! !!!! !! !! ! !! ! !!!! ! ! !! ! !!!! ! !!!! !! !! !!! ! !! !!! !!!!! !! !!! !! !!! ! !! !!! ! !!! !! !! ! !! !! !!! !! ! ! !!! ! !! ! !! !! !! ! !! !! !! ! !! !!! !! !! ! ! !! ! !!! !! !! !!! !! ! !! !!! !! !!! !! ! !!! ! !! !! ! !!! !! !!!! !! !!! !!! !!! !!!!! ! !! ! !!! !! ! !!! !! !! !!! ! !! ! !! !!! !! !!! ! !! !!!! !! ! ! ! !! ! !! ! !!!! ! !! !! !!!! !! !! !!! !! !! !! !!! ! !! ! !! !! ! !! ! !!! ! !! !!! !!! !! ! !! !! !! !!!! ! !! !! !! !! !!! !!

!!! !!! !! !! !

!!! ! !!!!

! !!!!!!

!

!!

! !!

!

!!

!

!

!!!

!

!

!

How to estimate local speeds?How to estimate local speeds? One way is from isochronesOne way is from isochrones

Data by M. Vander Linden Data by M. Vander Linden (919 sites) (919 sites)

Interpolation with GIS software Interpolation with GIS software

Legend! 2008_10_23a_fepre_context_ppnb9000CALyrBP Events

lakes

country

2008_10_23a

<VALUE>

<6.669 cal yr BP (>73 gen)

6.669 - 7.264 cal yr BP (73-54 gen)

7.265 - 7.859 cal yr BP (54-36 gen)

7.860 - 8.454 cal yr BP (36-17 gen)

>8.455 cal yr BP (<17 gen)

Page 6: Data and mathematical approaches  to the neolithic transition

6

Bocquet-Appel et al., Bocquet-Appel et al., J. Arch. Sci.J. Arch. Sci. (2009), Fig. 8 (2009), Fig. 8Another way is to fit a surface of C-14 dates

and estimate the gradient

Page 7: Data and mathematical approaches  to the neolithic transition

7

400 600 800 1000 1200 1400 1600 1800 2000 2200

0.4

0.6

0.8

1.0

1.2

1.4

1.6

SP

EE

D

T NEOL

No trend.

Is the effect too small?

France >47ºN

France <47ºN

Greece

Italy

Germany

Belgium

Yugoslavia

Page 8: Data and mathematical approaches  to the neolithic transition

8

400 600 800 1000 1200 1400 1600 1800 2000 2200

0.4

0.6

0.8

1.0

1.2

1.4

1.6

SP

EE

D

T NEOL

No trend.

Is the effect too small?

Page 9: Data and mathematical approaches  to the neolithic transition

9

1. Archaeological data: Speed versus neolithisation time→ no trend

2. Mathematical models: Do they predict that this effect is too small?

3. Archaeological data: Speed versus latitude

Plan of this talkPlan of this talk

Page 10: Data and mathematical approaches  to the neolithic transition

10

““Predator-prey” modelsPredator-prey” models

1. Ammerman and Cavalli-Sforza→ (1984)

2. Lotka-Volterra:

PNP

PPP

P

PNN

NNN

N

ppP

ppa

tp

ppP

ppa

tp

max

max

1

1

2

2

2

2

max

2

2

2

2

max

1

1

y

p

x

pDpp

pp

pat

p

y

p

x

pDpp

pp

pat

p

PPPPN

P

PPP

P

NNNPN

N

NNN

N

2

2

km

gatherers-hunter ofnumber km

farmers ofnumber

P

N

p

p

births-deaths

Page 11: Data and mathematical approaches  to the neolithic transition

11

3. Cohabitation models 2km

farmers ofnumber Np

before migration cohabitation non-cohabitation

(e.g., Lotka-Volterra)

pN time t gen time t + 1 gen time t + 1 gen

x x-r x x+r x-r x x+r

Example: if all individuals disperse:

Page 12: Data and mathematical approaches  to the neolithic transition

12

Cohabitation modelsCohabitation models

F et al, Phys Rev E (2007) F et al, New J Phys (2008) Isern et al, J Stat Mechs (2008)

2

2

km

gatherers-hunter ofnumber km

farmers ofnumber

P

N

p

p

fecunditynet

generation 1

),(

),(),(),(),(

0

0

N

N

PNNNN

R

T

rdtrx

trxPtrxPtrxPRTtxP

Dispersal probability distribution

Page 13: Data and mathematical approaches  to the neolithic transition

13

How to estimate the interaction parameter How to estimate the interaction parameter ΓΓ ? ?Effect of the interaction (no dispersal and Effect of the interaction (no dispersal and RR00NN=1)=1)

generation 1

)()()()(

T

tPtPtPTtPP PNNNN

)(max tPP PN

)(1

)()()(

)()( tPtPtPtP

tPtPP

NPN

P

PN

N

maxmax

1

NP

0 Γ 1/PN max 1/PN

Page 14: Data and mathematical approaches  to the neolithic transition

14

R0N = 2.2 (Birdsell 1957)

T = 32 yr (Stauder 1971) Currant & Excoffier (2005):

pe = 0.38 (Stauder 1971) Pmax P = 0.064 people/km2

r = 50 km (Stauder 1971) Pmax N =20 Pmax P=1.28people/km2

Speed of the neolithic frontSpeed of the neolithic front

km/yr 09.10 c

km/yr 10.11

maxmax c

P N

T

rpppRc eePN

)1))(cosh(1(2ln

0

min max0

This maximum difference of speeds is only 1% !!

This effect seems negligible. Reason: Pmax P << PmaxN

New J Phys (2008)

Page 15: Data and mathematical approaches  to the neolithic transition

15

1. Archaeological data: Speed versus neolithisation time

2. Mathematical models: The effect is too small

3. Archaeological data: Speed versus latitude

Plan of this talkPlan of this talk

Page 16: Data and mathematical approaches  to the neolithic transition

16

36 38 40 42 44 46 48 50 52

0.4

0.6

0.8

1.0

1.2

1.4

1.6

SPEED

SP

EE

D (

k/yr

)

LAT (º)

The same data versus latitudeItaly

GreeceYugoslavia

France>47ºFrance<47º

Germany

Belgium

Page 17: Data and mathematical approaches  to the neolithic transition

17

36 38 40 42 44 46 48 50 52

0.4

0.6

0.8

1.0

1.2

1.4

1.6

SPEED

SP

EE

D (

k/yr

)

LAT (º)

Is there

a trend?

Page 18: Data and mathematical approaches  to the neolithic transition

18

35 40 45 50 55 60

1

2

3

4

spe

ed

(km

/yr)

LATITUDE (º)

All grid nodes in Bocquet-Appel et al (2009)

But many nodes are on the sea...

Page 19: Data and mathematical approaches  to the neolithic transition

19

Bocquet-Appel et al., Bocquet-Appel et al., J. Arch. Sci.J. Arch. Sci. (2009), Fig. 8 (2009), Fig. 8

Is there a trend using all grid nodes?

Page 20: Data and mathematical approaches  to the neolithic transition

20

36 38 40 42 44 46 48 50 52 54 56

1

2

3

4

spe

ed

(km

/yr)

LAT

All grid nodes in 7 regions

But the

Mediterranean

spread is

faster

Page 21: Data and mathematical approaches  to the neolithic transition

21

Page 22: Data and mathematical approaches  to the neolithic transition

22

36 38 40 42 44 46 48 50 52 54 56

1

2

3

4sp

ee

d (

km/y

r)

LAT

All grid nodes in 7 regions

MEDITERRANEAN

Page 23: Data and mathematical approaches  to the neolithic transition

23

48 49 50 51 52 53 54

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.2sp

ee

d (

km/y

r)

LAT

All grid nodes in 7 regionswith latitude >=48ºN

Trend with latitude

Page 24: Data and mathematical approaches  to the neolithic transition

24

48 49 50 51 52 53 54

0.5

1.0

1.5

2.0

speed (km/yr)sp

ee

d (

km/y

r)

LAT

Germany

Trend also

in some

smaller regions

(but only at

latitude>47º)

Page 25: Data and mathematical approaches  to the neolithic transition

25

How to find the speed direction?How to find the speed direction?What should the surface of C14 arrival dates look like?

W

S

Nfast

slow

Near East

RussiaBritain

Portugal

E

t arrival

5000 cal yr BP

13000 cal yr BP

Page 26: Data and mathematical approaches  to the neolithic transition

26

How to find the speed direction?How to find the speed direction?The gradient is orthogonal to the level lines*

*N. Piskunov, Differential and Integral Calculus (Moscow, 1966), Sec. 8.15

y

x

t arrival surface

level line

level plane

Page 27: Data and mathematical approaches  to the neolithic transition

27

The speed is also orthogonal to the level lines of arrival time (isochrones):

How to find the speed direction?How to find the speed direction?

Speed vector 8500 CAL yr BP

8000 CAL yr BP

So the speed vector is parallel to the gradient

Page 28: Data and mathematical approaches  to the neolithic transition

28

GermanyGermany

6 8 10 12 14 16

48

50

52

54

56

LA

T (

º)

LON (º)

Gradients of dates surface (yr/km),as in Bocquet-Appel et al., JAS (2009), Fig. 8

6 7 8 9 10 11 12 13 14 1546

47

48

49

50

51

52

53

54

55

LA

T (

º)

LON (º)

Speed vectors (km/yr)

slowdownto the North

Page 29: Data and mathematical approaches  to the neolithic transition

29

1. Archaeological dataNo observed trend of speed versus neolithisation time

2. Mathematical models

The effect is too small because Pmax P << PmaxN

3. Archaeological dataLocal trends of speed versus latitude >47ºN. Cause?

ConclusionsConclusions

Page 30: Data and mathematical approaches  to the neolithic transition

30

Possible causesPossible causes

1. Reproduction and/or mobility decreases with increasing latitude NO DATA AVAILABLE

2. More density of hunter-gatherers to the North + resistance to farmers? DATA AVAILABLE?

3. Time needed for crop to adapt to colder climates to the North?

DATA AVAILABLE?

To model To model locallocal trends, the cause has to be known trends, the cause has to be known