Maize origins in Mexico and SW US revealed through ancestral reconstruction
1. Introgression and the origin of maize
in Mexico and the Southwest US
Jeffrey Ross-Ibarra
@jrossibarra • www.rilab.org
Dept. Plant Sciences • Center for Population Biology • Genome Center
University of California Davis
2. acknowledgements
Matt Hufford
(Iowa State)
Rute Fonseca
(Copenhagen)
Joost van Heerwaarden
(Wageningen)
Tom Gilbert (Copenhagen) John Doebley (Wisconsin)
6. Meyerowitz 1994 Current Biology
Duvick et al. 1999 US 6639132 B1
maize origins
Tripsacum extinct maize
7. Meyerowitz 1994 Current Biology
Duvick et al. 1999 US 6639132 B1
maize origins
Tripsacum F1 extinct maize
8. Meyerowitz 1994 Current Biology
Duvick et al. 1999 US 6639132 B1
maize origins
Tripsacum F1 extinct maize
9. Meyerowitz 1994 Current Biology
Duvick et al. 1999 US 6639132 B1
maize origins
F1 Tripsacum F1 extinct maize
10. single origin from teosinte
Matsuoka et al. 2002 PNAS
Piperno et al. 2009 PNAS
mays
parv
mex
11. paradox of maize ancestry
Fig. 2. A view of the enormous boulder that formed the Xihuatoxtla Shelter.
Matsuoka et al. 2002 PNAS
Piperno et al. 2009 PNAS
mays
parv
mex
1,400 mm annually with over 90% falling during the June-to-
October rainy season. The potential vegetation of the region is
tropical deciduous forest, seen today in remnants scattered
throughout the region. Vegetational reconstructions indicate
that a species-diverse seasonal tropical forest flourished there
B.P.), but the reworking of the rock shelter deposits in the early
20th century has left them thoroughly mixed over most of the
site. At a fourth site, the Temaxcalapa Shelter, we recovered a
large number of chipped stone artifacts from the eroded slope in
front of the shelter, including preceramic-aged spear points.We
SEE COMMENTARY
13. SNPs show highland similarity to teosinte
van Heerwaarden et al. 2011 PNAS
14. admixture explains patterns of diversity
mexicana parviglumis South/Caribbean West Highland
2500
2000
m
1500
1000
500
0
van Heerwaarden et al. 2011 PNAS
15. admixture explains patterns of diversity
H Genetic distance to parviglumis Heterozygosity
0.0 0.2 0.4 0.6 0.8 1.0
0.20 0.25 0.30 0.35 0.40 0.45
admixture proportion
F
van Heerwaarden et al. 2011 PNAS
0.0 0.2 0.4 0.6 0.8 1.0
0.25 0.26 0.27 0.28 0.29
admixture proportion
observed
observed
17. admixture along the genome
maize
mexicana
Mb
% population admixed
Chromosome 4
Hufford et al. 2013 PLoS Genetics
18. El Porvenir
Opopeo
Santa Clara
Nabogame
Puruandiro
Xochimilco
Tenango del Aire
San Pedro
Ixtlan
Allopatric
admixture localized, asymmetric
El Porvenir
Opopeo
Santa Clara
Nabogame
Puruandiro
Xochimilco
Tenango del Aire
San Pedro
Ixtlan
Allopatric
Hufford et al. 2013 PLoS Genetics
Pyhäjärvi et al. 2013 GBE
19. El Porvenir
Opopeo
Santa Clara
Nabogame
Puruandiro
Xochimilco
Tenango del Aire
San Pedro
Ixtlan
Allopatric
admixture localized, asymmetric
El Porvenir
Opopeo
Santa Clara
Nabogame
Puruandiro
Xochimilco
Tenango del Aire
San Pedro
Ixtlan
Allopatric
Inv4n
Hufford et al. 2013 PLoS Genetics
Pyhäjärvi et al. 2013 GBE
20. El Porvenir
Opopeo
Santa Clara
Nabogame
Puruandiro
Xochimilco
Tenango del Aire
San Pedro
Ixtlan
Allopatric
admixture localized, asymmetric
El Porvenir
Opopeo
Santa Clara
Nabogame
Puruandiro
Xochimilco
Tenango del Aire
San Pedro
Ixtlan
Allopatric
Inv4n
Hufford et al. 2013 PLoS Genetics
Pyhäjärvi et al. 2013 GBE
21. El Porvenir
Opopeo
Santa Clara
Nabogame
Puruandiro
Xochimilco
Tenango del Aire
San Pedro
Ixtlan
Allopatric
admixture localized, asymmetric
El Porvenir
Opopeo
Santa Clara
Nabogame
Puruandiro
Xochimilco
Tenango del Aire
San Pedro
Ixtlan
Allopatric
*
*
*
* *
Inv4n
Hufford et al. 2013 PLoS Genetics
Pyhäjärvi et al. 2013 GBE
22. introgressions overlap with mexicana QTL
Lauter et al. 2004 Genetics
Moose et al. 2004 Genetics
b1 in maize
Inv4n
Hufford et al. 2013 PLoS Genetics
23. introgressions show mexicana phenotypes
Introgression No Introgression
Hufford et al. 2013 PLoS Genetics
macrohairs
pigmentaton
27. Gene frequency
a solution: ancestral reconstruction
Nicholson et al 2002 J Roy Stat Soc. B
wild ancestor
1
0
28. origin
Gene frequency
a solution: ancestral reconstruction
Nicholson et al 2002 J Roy Stat Soc. B
wild ancestor
1
0
1
0
29. wild ancestor origin
derivative 1 derivative 2
Gene frequency
a solution: ancestral reconstruction
1
drift
1
drift
Nicholson et al 2002 J Roy Stat Soc. B
1
0
1
0
0
0
32. because the adaptive differences between highland and lowland
maize are profound (14, 29). In other crops, uncertainty about
ancestral allele frequen-cies.
frequencies observed in parviglumis
because the adaptive differences between highland and maize are profound (14, 29). In other crops, uncertainty 0.1 0.2 0.3 0.4
maize origins in lowland west Mexico
frequen-cies.
parviglumis
B A B C
Compared to wild teosinte Compared to ancestral maize
0.1 0.2 0.3 0.4 0.1 0.2 0.3 0.4
Lowland
Lowland
0.3 0.4 0.1 0.2 0.3 0.4
0.1 0.2 0.3 0.4
Genetic distance from ancestor
density
F F
F F
F
parameter F for 10 genetic groups with respect to (A) mexicana and (B) parviglumis. Only lowland included. van (Heerwaarden C) Drift of all et 10 al. genetic 2011 PNAS
groups with respect to inferred ancestral frequencies. Light blue represents between lowlands (<1,500 m, solid line) and highlands (>2,000 m).
C
genetic groups with respect to (A) mexicana and (B) parviglumis. Only lowland accessions of
genetic groups with respect to inferred ancestral frequencies. Light blue represents West
m, solid line) and highlands (>2,000 m).
33. maize origins in lowland west Mexico
1750m 1250m
van Heerwaarden et al. 2011 PNAS
34. how did maize arrive and adapt to SW US?
J. ROSS-IBARRA / UC DAVIS
2. Coastal route
1. Highland route
SW
Mexico
Fonseca et al. 2015 Nature Plants
Ronald Humeyestewa
35. ancient cobs to investigate chronology
Fonseca et al. 2015 Nature Plants
Hufford et al. 2012 Nature Genetics
36. DNA capture to enrich for subset of genome
J. ROSS-IBARRA / UC DAVIS
Fonseca et al. 2015 Nature Plants
37. ancient DNA data analysis: ANGSD
J. ROSS-IBARRA / UC DAVIS
error rate
Fonseca et al. 2015 Nature Plants
38. SW shares ancestry with highland Mexico
J. ROSS-IBARRA / UC DAVIS
6. Barplots of STRUCTURE q-matrices (2310 modern maize landraces and
teosinte) from k=3 to k=10 where k is the number of groups assigned. Each individual
Figure group is sorted from low to high elevation (top to bottom). The high elevation groups in
the SW, Mexico and Guatemala share ancestry with each other and to varying extents
parviglumis
mexicana
S. America
C. America
Mexico
SWUS
Fonseca et al. 2015 Nature Plants
US
39. allele counts support both hypotheses
SWUS highland coast teosinte
HUFFORD ET AL. 2012 PLOS GENETICS
VAN HEERWAARDEN ET AL 2011 PNAS
Fonseca et al. 2015 Nature Plants J. ROSS-IBARRA / UC DAVIS
40. allele counts support both hypotheses
SWUS highland coast teosinte
K=5
a b c
TMx25
TMx08
TPa06
TPa10
Ancient Mexico
Reventador
Chapalote
Highlands
Mex35
Mex25
Mex31
Mex30
Mex29
SW3K
SW2K
SW750
USA17
0.00 0.05 0.10 0.15 0.20
Teosintes
USA
Mexico
Central/South America
Chile
Z. m. mexicana
Z. m. parviglumis
Coastal
Drift parameter
USA
Mexico
SW3K
SW2K
SW750
USA17
-0.2 -0.1 0.0 0.1 0.2
D
Coastal
Highlands
Outgroup
SW
Coastal
Highlands
Outgroup
SW
HUFFORD ET AL. 2012 PLOS GENETICS
VAN HEERWAARDEN ET AL 2011 PNAS
Figure 1
Fonseca et al. 2015 Nature Plants J. ROSS-IBARRA / UC DAVIS
41. tree, admixture analyses support results
K=5
TMx25
TMx08
TPa06
TPa10
Ancient Mexico
Reventador
Chapalote
Highlands
Mex35
Mex25
Mex31
Mex30
Mex29
SW3K
SW2K
SW750
USA17
0.00 0.05 0.10 0.15 0.20
Teosintes
USA
Mexico
Central/South America
Chile
Z. m. mexicana
Z. m. parviglumis
Coastal
Drift parameter
USA
Coastal
Highlands
SW
0.1 0.2
b c
Figure 1
Fonseca et al. 2015 Nature Plants J. ROSS-IBARRA / UC DAVIS
42. ancient population genomics
J. ROSS-IBARRA / UC DAVIS
~2000 years b.p.
n=10
Changes in
shape and size
~750 years b.p.
n=12
Fonseca et al. 2015 Nature Plants
44. selection affects patterns of diversity
SFS
Proportion
SNPs
0.7
0.525
0.35
0.175
0
1 2 3 4
Individuals
Proportion
SNPs
0.7
0.525
0.35
0.175
0
1 2 3 4
Individuals
Taj D = 0
Taj D < 0
45. selection affects patterns of diversity
SFS
Proportion
SNPs
0.7
0.525
0.35
0.175
0
1 2 3 4
Individuals
Proportion
SNPs
0.7
0.525
0.35
0.175
0
1 2 3 4
Individuals
Taj D = 0
Taj D < 0
PBS
2000yr
700yr
teosinte
46. methods identify known domestication loci
J. ROSS-IBARRA / UC DAVIS
median PBS values
parviglumis
SW750 SW2K
dehyd1A
parviglumis
SW2K
SW750
a b
Fonseca et al. 2015 Nature Plants Figure 2
47. domestication loci
John Doebley
median PBS values parviglumis
parviglumis)
PBS(SW750 SW2K
dehyd1A
zagl1
su1
parviglumis
SW750 SW2K
parviglumis
parviglumis
SW2K
b
Fonseca et al. 2015 Nature Plants J. ROSS-IBARRA / UC DAVIS Liu et al. 2013 PLOS Genetics
48. domestication loci
John Doebley
median PBS values parviglumis
parviglumis)
PBS(SW750 SW2K
dehyd1A
zagl1
su1
parviglumis
SW750 SW2K
parviglumis
parviglumis
SW2K
b
Fonseca et al. 2015 Nature Plants J. ROSS-IBARRA / UC DAVIS Liu et al. 2013 PLOS Genetics
49. Fonseca et al. 2015 Nature Plants J. ROSS-IBARRA / UC DAVIS
PBS(parviglumis)
parviglumis
SW750 SW2K
dehyd1A
su1
parviglumis
SW750 SW2K
parviglumis
SW2K
SW750
parviglumis
SW2K
SW750
Figure 2
Liu et al. 2013 PLOS Genetics
median PBS values
PBS(parviglumis)
parviglumis
SW750 SW2K
dehyd1A
zagl1
su1
parviglumis
SW750 SW2K
parviglumis
parviglumis
SW2K
b
adaptation loci
Bill Belknap
John Doebley
50. chronology of selection on starch pathway
Fonseca et al. 2015 Nature Plants J. ROSS-IBARRA / UC DAVIS
51. concluding thoughts
• introgression confounds simple estimates of origin
• ancestral reconstruction provides resolution
• introgression adapted maize to highlands
• Southwest US maize originated in highlands with
subsequent gene flow from Pacific Coast
• ancient DNA allows chronology of gene flow,
adaptation