The Irish potato famine pathogen Phytophthora infestans originated in central Mexico rather than the Andes

Goss, E. M., Tabima, J. F., Cooke, D. E. L., Restrepo, S., Fry, W. E., Forbes, G. A., ... & Grünwald, N. J. (2014). The Irish potato famine pathogen Phytophthora infestans originated in central Mexico rather than the Andes. Proceedings of the National Academy of Sciences, 111(24). doi:10.1073/pnas.1401884111

10.1073/pnas.1401884111

National Academy of Sciences

Version of Record

http://cdss.library.oregonstate.edu/sa-termsofuse

The Irish potato famine pathogen Phytophthorainfestans originated in central Mexico rather thanthe AndesErica M. Gossa, Javier F. Tabimab, David E. L. Cookec, Silvia Restrepod, William E. Frye, Gregory A. Forbesf,Valerie J. Fielandb, Martha Cardenasd, and Niklaus J. Grünwaldg,h,1

aDepartment of Plant Pathology and Emerging Pathogens Institute, University of Florida, Gainesville, FL 32611; bDepartment of Botany and Plant Pathology,Oregon State University, Corvallis, OR 97331; cThe James Hutton Institute, Invergowrie, Dundee DD2 5DA, Scotland; dDepartment of Biological Sciences,University of the Andes, 110321 Bogota, Colombia; eDepartment of Plant Pathology and Plant-Microbe Biology, Cornell University, Ithaca, NY 14853; fCIPChina Center for Asia and the Pacific, International Potato Center, Beijing 100081, China; gHorticultural Crops Research Laboratory, US Department ofAgriculture Agricultural Research Service, Corvallis, OR 97330; and hDepartment of Botany and Plant Pathology and Center for Genome Biology andBiocomputing, Oregon State University, Corvallis, OR 97331

Edited by Detlef Weigel, Max Planck Institute for Developmental Biology, Tübingen, Germany, and approved May 6, 2014 (received for reviewJanuary 30, 2014)

Phytophthora infestans is a destructive plant pathogen best knownfor causing the disease that triggered the Irish potato famine andremains the most costly potato pathogen to manage worldwide.Identification of P. infestan’s elusive center of origin is critical tounderstanding the mechanisms of repeated global emergence ofthis pathogen. There are two competing theories, placing the originin either South America or in central Mexico, both of which arecenters of diversity of Solanum host plants. To test these competinghypotheses, we conducted detailed phylogeographic and approxi-mate Bayesian computation analyses, which are suitable approachesto unraveling complex demographic histories. Our analyses usedmicrosatellite markers and sequences of four nuclear genes sampledfrom populations in the Andes, Mexico, and elsewhere. To infer theancestral state, we included the closest known relatives Phytoph-thora phaseoli, Phytophthora mirabilis, and Phytophthora ipo-moeae, as well as the interspecific hybrid Phytophthora andina. Wedid not find support for an Andean origin of P. infestans; rather, thesequence data suggest a Mexican origin. Our findings support thehypothesis that populations found in the Andes are descendantsof the Mexican populations and reconcile previous findings of an-cestral variation in the Andes. Although centers of origin are welldocumented as centers of evolution and diversity for numerous cropplants, the number of plant pathogens with a known geographicorigin are limited. This work has important implications for our un-derstanding of the coevolution of hosts and pathogens, as well asthe harnessing of plant disease resistance to manage late blight.

biological invasion | coalescent analysis | oomycete | population genetics |stramenopile

The potato pathogen Phytophthora infestans, the causal agentof potato late blight, is the plant pathogen that has most

greatly impacted humanity to date. This pathogen is best knownfor its causal involvement in the Irish potato famine after in-troduction of the HERB-1 strain to Ireland from the Americasin the 19th century (1). To this day, potato late blight remainsa major threat to food security and carries a global cost con-servatively estimated at more than $6 billion per year (2). Inthe 1980s, a single asexual lineage named US-1, possibly derivedfrom the same metapopulation as HERB-1 (1), dominated glob-al populations, whereas a genetically diverse and sexual popula-tion of P. infestans in central Mexico led to formulation of the hy-pothesis identifying Mexico as this pathogen’s center of origin (3,4). A competing hypothesis argues that the center of origin of thepotato, the South American Andes, is the center of origin ofP. infestans (5). This hypothesis recently gained prominence afteran analysis demonstrated ancestral variation in Andean lineagesof P. infestans (5). Other evidence supporting this hypothesisincludes infection of native Solanum hosts and an Andean

distribution for Phytophthora andina, a phylogenetic relative of P.infestans (6).Evidence supporting a Mexican center of origin is substantial,

but inconclusive (4). Two close relatives of P. infestans, Phy-tophthora ipomoeae and Phytophthora mirabilis, are endemic tocentral Mexico (7, 8). P. ipomoeae and P. mirabilis cause diseaseon two endemic plant host groups, Ipomoea spp. and Mirabilisjalapa, respectively. Populations of P. infestans in the TolucaValley, southwest of Mexico City, are genetically diverse, are inHardy–Weinberg equilibrium, and contain mating types A1 andA2 in the expected 1:1 ratio for sexual populations (9, 10). Be-fore a migration event from Mexico to Europe in the 1970s (11,12), only A1 mating types of P. infestans were found worldwideoutside of central Mexico, limiting other populations to asexualreproduction (13). Tuber-bearing native Solanum species occurthroughout the Toluca Valley (14). Of the R genes that havebeen used to confer resistance to strains of P. infestans in potato,the majority described to date originated from Solanum demis-sum or Solanum edinense in the Toluca Valley, with some dis-covered in South America (15).

Significance

The potato late blight pathogen was introduced to Europe inthe 1840s and caused the devastating loss of a staple crop,resulting in the Irish potato famine and subsequent diaspora.Research on this disease has engendered much debate, whichin recent years has focused on whether the geographic originof the pathogen is South America or central Mexico. Differentlines of evidence support each hypothesis. We sequenced fournuclear genes in representative samples from Mexico and theSouth American Andes. An Andean origin of P. infestans doesnot receive support from detailed analyses of Andean andMexican populations. This is one of a few examples of a path-ogen with a known origin that is secondary to its currentmajor host.

Author contributions: E.M.G., D.E.L.C., S.R., G.A.F., and N.J.G. designed research; E.M.G.,J.F.T., D.E.L.C., S.R., W.E.F., G.A.F., V.J.F., M.C., and N.J.G. performed research; D.E.L.C., S.R.,W.E.F., G.A.F., and N.J.G. contributed new reagents/analytic tools; E.M.G., J.F.T., D.E.L.C.,V.J.F., M.C., and N.J.G. analyzed data; and E.M.G., J.F.T., D.E.L.C., S.R., W.E.F., G.A.F., M.C.,and N.J.G. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: The sequences reported in this paper have been deposited in the GenBankdatabase (accession nos. KF979339–KF980878).1To whom correspondence should be addressed. E-mail: grunwaln@science.oregonstate.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1401884111/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1401884111 PNAS | June 17, 2014 | vol. 111 | no. 24 | 8791–8796

AGRICU

LTURA

LSC

IENCE

S

Support for the alternate hypothesis that P. infestans origi-nated in the Andes is based on a coalescent analysis conductedby Gómez-Alpizar et al. (5). This analysis used the nuclear RASlocus and the mitochondrial P3 and P4 regions to infer rootedgene genealogies that showed ancestral lineages rooted in theAndes. Furthermore, the Mexico sample harbored less nucleo-tide diversity than the Andean population. P. andina was iden-tified as the ancestral lineage for the mitochondrial genealogy;however, P. mirabilis and P. ipomoeae were not included in thatstudy. P. andina has since been shown to be a hybrid species derivedfrom P. infestans and a Phytophthora sp. unknown to science (16).Surprisingly, populations of P. infestans and P. andina are clonal inSouth America and are not in Hardy–Weinberg equilibrium (6, 17–19). Thus, the question of whether P. infestans originated in theAndes or central Mexico remained unresolved.Powerful approaches for determining the demographic and

evolutionary history of organisms are now available (20). Manyof these approaches rely on the power of coalescent theory forinferring the genealogical history of a species based on a repre-sentative population sample (21–23). Bayesian phylogeographyuses geographic information in light of phylogenetic uncertaintyto provide model-based inference of geographic locations ofancestral strains (24). The isolation with migration (IM) modeland associated software uses likelihood-based inference to inferdivergence time between evolutionary lineages (25). ApproximateBayesian computation (ABC) makes use of coalescent simulationsand likelihood-free inference to contrast complex demographicscenarios. Each of these methods has proven useful in recon-structing the demography of pests and pathogens (24, 26–29).The objective of the present study was to reconcile the two

competing hypotheses on the origin of P. infestans using Bayesianphylogenetics and ABC. We sampled key populations of P. infestansfrom central Mexico and the Andes and expanded on the anal-ysis of Gómez-Alpizar et al. (5) by sequencing additional nuclearloci to assess support for the center of origin across multiple loci.To determine ancestral state, we added sequences from the sistertaxa P. andina, P. mirabilis, P. ipomoeae, and Phytophthora phaseoli,all of which belong to Phytophthora clade 1c (30, 31). Finally, weaimed to reconcile the biology of P. infestans in Mexico with thefindings of Gómez-Alpizar et al. (5) of ancestral variation inthe Andes.

ResultsPopulation Structure and Mode of Reproduction. Our sample ofP. infestans included 40 isolates from Colombia, Ecuador, andPeru and 48 isolates from Toluca Valley and Tlaxcala State incentral Mexico (SI Appendix, Table S1). We found 30 multilocussimple sequence repeat (SSR) genotypes in the Andes sampleand 43 in the Mexico sample. The Mexico sample had slightlyhigher mean allelic richness per locus compared with the Andessample (6.7 vs. 5.2) after correction for sample size by rarefac-tion. The mean number of private alleles per locus for the Mexicosample was more than twice that for the Andes sample (2.1 vs.0.8). In the Mexico sample, clonality was detected in isolatessampled from single patches of the indigenous host S. demissum.The index of association, IA, calculated for clone-corrected datafor the Mexico isolates accepted the hypothesis of sexual re-production (P = 0.25). In contrast, the hypothesis of no linkageamong markers was rejected for the Andes sample (P < 0.001),supporting a clonal mode of reproduction.We inferred population structure based on SSR genotypes

separately for the Mexico and Andes samples of P. infestans,based on a priori knowledge of their sexual and clonal repro-duction, respectively, using structure (32). The Mexico sampleconsisted of admixed subpopulations with at least K = 4 under-lying groups (Fig. 1A and SI Appendix, Figs. S1 and S2), whereasthe Andes sample consisted of two distinct clusters with verylittle admixture (Fig. 1B and SI Appendix, Figs. S1 and S2).

Phylogeographic Root of P. infestans.We used Bayesian multilocusphylogeographic analysis to infer the geographic location of the

root (“root state”) of P. infestans and the Phytophthora clade 1cspecies using BEAST (24, 33). For this analysis, we includeda representative global sample including isolates from the now-diverse populations in Europe (SI Appendix, Table S1). Com-parison of different molecular clocks for each sequenced nuclearlocus showed that three of four loci fit a strict clock, in which thestandard deviation of the uncorrelated lognormal relaxed mo-lecular clock indicated no variation in rates among branches. Incontrast, the RAS locus (intron Ras and Ras fragments) showedhigh variation in rates among branches and required a relaxedlognormal molecular clock (SI Appendix, Table S2). The PITG_11126locus had the highest rate of evolution, whereas β-tubulin (b-tub)had the lowest substitution rate.Root state reconstruction produced the highest posterior prob-

abilities for Mexico as the root state of both the P. infestans andclade 1c datasets (Fig. 2). For each locus independently, posteriorprobabilities of a Mexico root were >0.8 for clade 1c (SI Appendix,Fig. S3). In nearly all of the P. infestans clades, there was aninferred ancestral connection to Mexico (SI Appendix, Figs. S4–S8). All of the species in Clade 1c were monophyletic with highsupport, except for the hybrid species P. andina as previouslydemonstrated (16, 34).

Evolution of P. infestans in the Andes. We found that three out offour loci had either greater nucleotide diversity or Watterson’stheta for the Andes sample compared with the Mexico sample(SI Appendix, Table S3). To better understand the evolution ofthe Andes lineages and the relationship between P. infestans inMexico and the Andes, we estimated pairwise times since di-vergence using combined SSR genotypes and nuclear sequencedata using the IMa program (25). Divergence times were esti-mated between each of the clonal lineages in the Andes sample(US-1, EC-1, PE-3, and PE-7) and one another and the Mexicosample (Fig. 3). EC-1, PE-3, and PE-7 produced recent pairwisedivergence times. EC-1 showed the most recent divergence fromUS-1 and the Mexico population. The time since divergence ofPE-3 and US-1 was less than that of PE-3 and Mexico. The timesince divergence between PE-7 and Mexico was greater than forEC-1 or PE-3 from Mexico. The marginal posterior probabilitydistribution for the divergence time between PE-7 and US-1 wasflat, indicating uncertainty in the history of PE-7.We further explored scenarios for the evolution of the Andes

lineages by testing a series of alternative models using ABC asimplemented in the DIYABC program (35). We first examined

0 1 2

50

100

150

200

Index of association

Cou

nt

50

100

150

Cou

nt

0-0.2-0.4 0.2 0.4Index of association

B

A

PiM

X29

P

iMX

65

PiM

X21

P

iMX

22

PiM

X02

P

iMX

15

PiM

X23

P

iMX

04

PiM

X20

P

iMX

03

PiM

X12

Pi

MX4

1 P

iMX

61

PiM

X62

P

iMX

27

PiM

X40

PiM

X14

P

iMX

16

PiM

X10

Pi

MX4

2 P

iMX

06

PiM

X26

P

iMX

18

PiM

X30

P

iMX

19

PiM

X11

PiM

X24

P

iMX

25

PiM

X28

P

iMX

01

PiM

X64

P

iMX

67

PiM

X63

P

iMX

66

PiM

X13

P

iMX

17

PiM

X72

P

iMX

71

PiM

X05

P

iMX

07

PiM

X50

PiM

X48

PiM

X49

PiM

X43

PiM

X44

PiM

X45

PiM

X46

PiM

X47

ObservedIA = 0.08P = 0.25

ObservedIA = 2.63P < 0.001

EC1 PE7 PE3 US1US8

PiP

E01

P

iCO

01

PiC

O02

P

iCO

03

PiC

O04

P

iEC

03

PiP

E11

PiP

E12

P

iPE

14

PiP

E05

P

iPE

09

PiE

C01

P

iPE

04

PiP

E02

P

iPE

08

PiP

E13

P

iPE

07

PiP

E26

P

iEC

02

PiP

E10

P

iPE

03

PiP

E06

P

iCO

05

PiP

E21

P

iEC

06

PiE

C04

P

iPE

20

PiE

C11

PiE

C12

P

iEC

07

PiE

C08

P

iEC

10

PiE

C13

P

iEC

14

PiP

E27

P

iEC

05

PiP

E22

P

iPE

23

PiP

E25

P

iPE

24

Andes

Mexico

Fig. 1. Population structure of Mexico and Andes samples of P. infestansinferred using structure (32). (A) The Mexico sample shows admixed indi-viduals assigned to K = 4 clusters. Isolates collected from patches ofS. demissum are in bold type. Based on the index of association, IA, there isno evidence of linkage disequilibrium among loci (P = 0.25), consistent witha sexually recombining population. (B) The Andes sample clusters into K = 2distinct clades with little or no admixture. The hypothesis of no linkageamong markers is rejected (P < 0.001), indicating a clonal population.

8792 | www.pnas.org/cgi/doi/10.1073/pnas.1401884111 Goss et al.

the relationships among the lineages EC-1, PE-3, and PE-7.Model comparison indicated that the PE-3 and PE-7 lineageshave more recent shared ancestry compared with the morewidespread EC-1 lineage (SI Appendix, Fig. S9). We next testedthe relationships of each of these lineages to the Toluca pop-ulation in Mexico and the US-1 lineage in the Andes (SI Ap-pendix, Fig. S10). Here we used only isolates from the TolucaValley, because we know that this population is panmictic (4).Preliminary analyses showed support for the PE-3 and PE-7lineages being both closely related and distantly related to theother lineages.To better understand this pattern, we included scenarios

containing all possible pairwise admixture events among pop-ulations as well as an admixture with an unsampled population(36) (SI Appendix, Fig. S10). There was relatively strong supportfor PE-3 as an admixed lineage. The scenario in which PE-3 isderived from an admixture event between US-1 and an unsam-pled population had a posterior probability of 0.67 (Fig. 4A andSI Appendix, Table S4). All other scenarios had posterior prob-abilities <0.14, most <0.01. The equivalent scenario for PE-7also had the highest posterior probability of the tested modelsat 0.37, but there was also support for PE-7 emerging froman admixture event between the Toluca population and theunsampled population (Fig. 4 B and C and SI Appendix, Table S4).The relationship of EC-1 to US-1 and the Toluca population wasuncertain as well, with nearly equal support for two scenarios (Fig.4 D and E and SI Appendix, Table S4). In one scenario, the EC-1lineage recently diverged from the Toluca population (P = 0.25),and in the other, EC-1 was an admixture of the Toluca populationand an unsampled population (P = 0.31). Our Andes samples werehighly clonal; thus, we interpret these results as indicating that thePE-3, PE-7, and perhaps EC-1 lineages formed after a sexual eventbetween two distinct lineages or populations.We used the most highly supported scenarios to test more

complex scenarios including all four Andean lineages andToluca. Testing all possible admixture events proved to be toocomplex with support split among scenarios with various com-binations of admixture events. Furthermore, the PE-7 resultsfrom both IMa and DIYABC analyses suggest that this lineagehas a complex ancestry. Therefore, the final scenarios testedadmixture in the evolution of EC-1 and PE-3, and excluded PE-7(Fig. 5). We found that scenario B, in which EC-1 split from theToluca population and PE-3 originated from an admixture event,had the highest posterior probability of 0.74 (Fig. 4). Notably,there was minimal support for scenario C, with simple ancestraldivergence of PE-3 (Fig. 5).

We estimated type I and type II errors for scenario B andfound that 65% of the datasets simulated under scenario Bproduced the highest posterior probability for scenario B (type Ierror, 0.352). Pseudo-observed datasets generated under scenariosA and C were wrongly assigned to scenario B at frequencies of0.210 and 0.096, respectively (type II error). Posterior distributionsfor parameters were wide, indicating limited confidence in theparameter estimates.

DiscussionWe found multilocus support for a Mexican origin of P. infestans.Bayesian phylogeographic analysis rooted both P. infestans andPhytophthora clade 1c in Mexico for each of four nuclear loci. Ourresults are consistent with the population biology of P. infestans incentral Mexico and, taken together, point to Mexico as the originof this pathogen. These results are supported by the previouslynoted pathogen and host characteristics. Specifically, Tolucapopulations are sexual, whereas South American populationsof P. infestans are clonal (6, 9, 10, 17–19). Both mating types ofP. infestans are known to have been present since at least the1960s in central Mexico (37, 38).The genealogical connections between continents that we

observed in our phylogeographic analysis are consistent with themovement of P. infestans among widespread potato growingregions. Migration estimates using microsatellite variation alsosupport our sequence analysis by showing migration fromMexicoto the Andes, but not from the Andes to Mexico (SI Appendix).The commercial potato seed trade can explain much of thecurrent global population structure of P. infestans; however, theearly movements of P. infestans have not been fully reconstructed(1), and examination of the processes underlying the emergenceof new, highly successful strains is ongoing (3, 39). The diversepopulation in central Mexico may be the ultimate source for theappearance of new strains worldwide (3, 40), although seed po-tatoes from Europe are behind recent migrations of virulent strains(41, 42). Detailed genetic reconstruction of global migrations ofP. infestans may be feasible using population genomic data.We explored the evolution of the Andean lineages and the

relationship between P. infestans in Mexico and the Andes bytesting a series of alternative models using the ABC technique.Our intention was to determine the timing of divergence of theAndean lineages and their relationship to the Toluca populationin Mexico. Surprisingly, we found evidence for diversification ofthe Andes population as a result of admixture or hybridization.Because of our limited power to discriminate between models,we view this analysis as hypothesis-generating. Nevertheless,even moderate support for hybridization generating novel

B

CCO EC ES HU MX NL PE SA SW UK US VT

CO EC ES HU MX NL PE PO SA SWUK US VT

P. infestans

Roo

t sta

te p

oste

rior p

roba

biliy

0.00

0.05

0.10

0.15

0.20

0.0

0.1

0.2

0.3

0.4

0.5

Roo

t sta

te p

oste

rior p

roba

biliy

MX

EC

MX

MX

MX

MX

P. ipomoeae

P. mirabilis

P. andina

P. infestansP. andina

LocationMexicoEcuador

1.0

0.9

0.8

1.0

1.0

1.0

1.0

MX

A Clade 1c

Fig. 2. Root state probabilities by country inferred using BEAST (33). (A)Summarized maximum clade credibility phylogeny of Phytophthora clade 1cspecies. Colors of branches indicate the most probable geographic origin ofeach lineage. Posterior probabilities of major branches supporting the treetopology are shown below each branch. Root state posterior probabili-ties indicate that Mexico is the most probable origin of clade 1c (B) andP. infestans (C ).

0

50

0

50

0

50 0

50

Scaled time since divergence

Est

imat

ed m

argi

nal p

oste

rior p

roba

bilit

y

Mexico

US-1

PE-7

PE-3

PE-3 PE-7EC-1 US-1

2.002.002.00 0 0.2

Fig. 3. Estimated marginal posterior probabilities of relative divergencetimes in pairwise comparisons among Andes clonal lineages (EC-1, PE-3, PE-7,and US-1) and the Mexico population estimated using IMa (64). Smallermeans and modes of the scaled time since divergence indicate more recentdivergence of lineages. Lineages EC-1, PE-3, and PE-7 show more recent di-vergence from one another than from US-1 and Mexico.

Goss et al. PNAS | June 17, 2014 | vol. 111 | no. 24 | 8793

AGRICU

LTURA

LSC

IENCE

S

lineages of P. infestans is compelling, given that a new pathogenof Solanum hosts, P. andina, has been generated via hybridiza-tion (16), and that infrequent hybridization among lineages ofP. infestans is suspected to be responsible for novel genotypeselsewhere (43).Considering the diversity of Solanum hosts in the Andes, there

is great potential for clonal diversification of P. infestans asavailable niches are colonized and rare events contribute togeneration or recombination of variation. There is no evidenceof genetic variation among P. infestans isolates from potatoes inPeru as recently as the mid-1980s (44). If there were diversity inP. infestans or clade 1c in the Andes, it must have been limited toother unsampled hosts. Based on our analyses, we hypothesizethat the Andean diversity in P. infestans has been driven by globalmigration together with hybridization among populations estab-lished via independent migration events.Given our results, why did Gómez-Alpizar et al. (5) infer an

Andean origin? First, their mitochondrial coalescent gene treeincluded the Andean endemic P. andina, but not the Mexican sisterspecies, and thus species selection rooted the tree in the Andes. Atthe time that the work was conducted, P. andina was not recog-nized as a hybrid species with two haplotypes derived from twodistinct parental species (16). When we removed P. andina fromtheir dataset, the location of the root was ambiguous (SI Appendix);therefore, the mitochondrial loci used in the analysis were notphylogeographically informative. The root of the P. infestans mi-tochondrial genome was recently dated to 460 y ago (95% highest-probability density, 300–643), around the time of the Spanishconquest of the Americas (1). Thus, the two major mitochondrialhaplotypes may be the product of movement of the pathogenby humans, resulting in the formation of a new population ofP. infestans and the evolution of diverged mtDNA haplotypesbefore global expansion of the pathogen some 200 y later.Second, the coalescent root of the single nuclear locus used by

Gómez-Alpizar et al. is dependent on migration rate (SI Appendix),which was estimated using the same locus. The RAS gene inP. infestans has two diverged haplotypes in the first intron. Thiscreates two diverged clades of haplotypes, one of which is not foundin the Mexican sample. This might have biased the migration rateestimates for this gene and affected the outcome of the coalescentanalysis. Our analysis of this locus required a relaxed molecularclock and took an exceptionally long time to converge. Finally,

explicit treatment of geography in our BEAST analysis did notrequire us to make assumptions about population structure, butrather incorporated the divergent origins of our isolates intothe analysis.Resolving the origin of P. infestans is important to our un-

derstanding of the emergence and reemergence of damagingpathogens. P. infestans is one of a limited number of agriculturalplant pathogens with a well-characterized center of origin (45).An expectation of long-term coevolution between a crop and itshost-specific pathogen can mislead one into thinking that thepathogen originates from the crop’s center of origin; however,there are other well-documented and suspected instances ofhost-jumping by crop pathogens (45–47). P. infestans appears tobe an example of a pathogen originating from wild relatives ina secondary center of host diversity. Identification of the centerof origin also advances our understanding of pathogen evolutionoutside of this region, that is, how the pathogen has changedafter introduction to new environments.Global populations of P. infestans have undergone rapid

changes owing to both migration and evolution after migration(39, 48). Knowledge of pathogen diversity and evolution bothwithin and outside of the center of origin is critical to cropbreeding efforts (49). The long-term success of efforts to breedlate blight-resistant potatoes will require breeders to accountfor geographic and evolutionary sources of novel variation inP. infestans and its hosts.

Materials and MethodsSampling Individuals. A sampling of the known global diversity in P. infestanswas obtained from several collaborators (SI Appendix, Table S1). We focusedon assembling two large samples from Mexico (n = 48) and the SouthAmerican Andes (n = 40) representing the known diversity from each ofthese regions, and included a more moderate global sample of P. infestansfor context (SI Appendix, Table S1).

SSR Typing. Isolates were genotyped using 11 SSR markers as describedpreviously (39, 50). Given that the individuals varied in ploidy, analysis wasrestricted to approaches that can accommodate nondiploid genetics (51).

Population Structure (SSR). Allelic richness was calculated using the ADZEprogram (52). Population composition was inferred using structure 2.3 (32)by testing the number of population clusters (K ) between 1 and 20 usingthe admixture model for the Mexico and Andes samples (53) and the no-admixture model for the Andes sample. The no-admixture model may bemore appropriate for the Andean population (i.e., Ecuador, Colombia, andPeru) given a priori knowledge of its clonality (17). Analysis was performedseparately for the two different populations. A total of 10 independentruns each of 1,000,000 iterations with a burn-in period of 20,000 Markovchain Monte Carlo (MCMC) iterations were conducted. The results fromstructure were postprocessed using Structure Harvester (54). The ΔKmethod was used to evaluate the rate of change in the log probability ofdata between successive K values, to infer the number of clusters (55).

Toluca US-1EC-1

DP = 0.25

Toluca US-1 PE-7

BP = 0.37

US-1 Toluca PE-7

CP = 0.29

Toluca US-1 PE-3

AP = 0.67

US-1 Toluca EC-1

EP = 0.31 Time

Fig. 4. Scenarios for the evolution of P. infestans in the Andes tested usingABC in DIYABC (35). Shown are the scenarios with the highest posteriorprobabilities out of 15 scenarios testing the relationships of PE-3, PE-7, andEC-1 lineages, in turn, to the US-1 lineage and Toluca Valley population (SIAppendix, Fig. S10). Present-day populations are at the base of the treeschematic. Ancestral relationships among these populations are representedby lines intersecting in the past, with the vertex of the schematic repre-senting the most recent common ancestor of all samples. Horizontal linesindicate admixture events between the ancestral populations connected bythe horizontal line. Potential changes in population size over time are in-dicated by changes in line thickness, but line thickness is not proportional topopulation size. The dashed line represents an unsampled population thathas contributed to the genetic variation observed in sampled populations.(A) For PE-3, the most probable scenario includes an admixture of US-1 andan unsampled population, leading to the PE-3 lineage. (B and C) For PE-7,support is split between two scenarios containing an admixture event withan unsampled population. (D and E) For EC-1, the scenarios with the greatestsupport showed a simple divergence from the Toluca population or an ad-mixture between Toluca and an unsampled population. Posterior probabil-ities for all tested scenarios and their 95% confidence intervals are given in SIAppendix, Table S4.

CBATime

Toluca EC-1 US-1 PE-3Toluca EC-1 US-1 PE-3 Toluca EC-1 US-1 PE-3

P = 0.02 [0.016, 0.028]

P = 0.74[0.686, 0.791]

P = 0.24[0.187, 0.292]

Fig. 5. Final three scenarios used to examine the relationships between theToluca Valley population and EC-1, PE-3, and US-1 lineages in the Andes byABC. Tree schematics are drawn as in Fig. 4. The three scenarios are simpledivergence of populations such that PE-3 is an ancestral lineage (A); emer-gence of the PE-3 lineage after an admixture of US-1 and an unsampledpopulation (B); and scenario B plus emergence of the EC-1 by an admixturebetween the Toluca population and an unsampled population (C). ScenarioB was the most likely of the three, with a posterior probability of 0.74. The95% confidence intervals for the posterior probabilities are in brackets.

8794 | www.pnas.org/cgi/doi/10.1073/pnas.1401884111 Goss et al.

The mode of reproduction was assessed by evaluating observed linkageamong loci against expected distributions from permutation using the indexof association, IA, as proposed by Brown et al. (56) and applied in multilocus(57). IA was calculated using the poppr package (58) in R (59) separately forthe Mexico and Andes samples and evaluated with 2,000 permutations usingclone-corrected data.

DNA Amplification and Sequencing. Nuclear loci were amplified and directlysequenced as described by Goss et al. (16). A larger fragment of the b-tub genewas sequenced than that sequenced by Gómez-Alpizar et al. (5). Haplotypephases of nuclear gene sequences with multiple heterozygous sites wereinferred using PHASE software (51). Inferred haplotypes were confirmed bycloning PCR products for a subset of genotypes and sequencing inserts. Inseveral instances, three distinct alleles were recovered from a given individual;these were the only times when the PHASE-inferred haplotypes were notvalidated. In some instances, isolates with three alleles at one locus werecloned at another locus, and only two alleles were found; thus, it was notassumed that these isolates had three distinct alleles at other loci.

For the phylogenetic and coalescent-based analyses, we removed indels andrecombinant alleles from the datasets. We used the four-gamete test to detectsignals of recombination, and found recombination in the b-tub, PITG_11126,and RAS alignments. For PITG_11126, the last 65 nucleotides were excludedowing to recombination in this region relative to the remainder of the frag-ment. For b-tub, the signal of recombination was removed when an indel wasdeleted from the alignment and isolate PiEC01was excluded. For the RAS locus,PiEC07, PiUS11, and PiUS17 were excluded.

The sequences generated have been deposited in GenBank (accession nos.KF979339–KF980878).

Demographic and Phylogeographic History.We adopted a Bayesian coalescentapproach to investigate divergence and phylogeographic history using BEAST1.7.4 (33, 60). BEAST implements a method for sampling all trees that havea reasonable probability given the data. The analysis is based on haplotypes,and two haplotypes in a given individual may or may not share a mostcommon recent ancestor. Furthermore, genealogies obtained from BEASTnecessarily estimate how old the common ancestor of these haplotypesmight be for each haplotype (even when haplotype sequences are identical),resulting in slight branch length variation among identical haplotypes. Thisis an expected outcome of coalescent analysis in which identical sequencesat a given locus are expected to show divergence at the genome level. Notethat although branch length variation is observed among identical hap-lotypes, there is no support for nodes at this level until different haplotypesequences coalesce to their common recent ancestors, and this branchlength variation should not be interpreted.

To initialize BEAST, we obtained the most likely models of nucleotidesubstitution for each alignment using jModelTest 2.1 (61) and ModelGenerator0.57 (62), and selected the consensus model from these programs (SI Appendix,Table S5). To ensure adequate sampling of parameters, we conducted theBEAST analyses with independent MCMC simulations of 200 million iter-ations for each locus both for P. infestans only and with the other clade 1cspecies. Analyses were conducted at least twice to ensure repeatability. Foreach MCMC run, we sampled every 10,000 generations and discarded non-stationary samples after 25% burn-in. Effective sample size estimates weretypically greater than 200, and parameter trace plots supported MCMCconvergence and good mixing in each independent run. Multilocus analyseswere also run for P. infestans only and also for all Phytophthora clade 1cspecies. For each dataset, models of nucleotide substitution, constant co-alescent priors (for the P. infestans dataset) or Yule speciation models (forthe clade 1c dataset) were used as specified priors to improve the calculationof clock rates, geographic ancestral states, and phylogenetic relationships.

To infer themost likely geographic origin for each clade, we used the discretephylogeographic approach as implemented by Lemey et al. (24). This methoduses the geographic locations of the samples to reconstruct the ancestral statesof tree nodes, including estimation of the root state posterior probability.Maximum clade credibility (MCC) trees were obtained using TreeAnnotator andvisualized in FigTree 1.3.1. The bar plots for the posterior probability of theroot state were created with the ggplot2 package (63) using R statisticalcomputing and graphic language (59). Maximum clade credibility trees wereobtained for each locus as well as for multigene analyses.

Divergence Times Under Isolation with Migration. IMa version 2.0 was used toestimate divergence times for each pair of Andean lineages and betweenAndean lineages and the Mexican population (64). The data used were the 4nuclear loci and 8 of the 11 SSR loci. Two SSR loci that did not conform to thestepwise mutation model (D13 and G11) and one SSR locus that was nearly

monomorphic (Pi33) were excluded. The infinite sites model was used for allnuclear loci. Initial maxima for uniform prior distributions of the parameterswere as follows: population size (q), 15.0; divergence time (t), 0.5; migrationrate (m), 1.0. Runs including the Mexican sample were also performed usinglarger priors for q (20.0) and t (1.0), and similar results were obtained. Theoption of running the burn-in and recording periods for indefinite durationswas chosen to ensure stabilization of likelihoods before the end of the burn-in and adequate sampling of the posterior distribution. Metropolis couplingwas implemented using the geometric increment model and 80 chains withgeometrical increment parameters of 0.97 and 0.3.

ABC. We evaluated alternative scenarios for the evolution of the AndeanP. infestans lineages in a systematic stepwise manner using ABC withDIYABC version 1.0.4.45 (35). ABC has been used to estimate parametersfor complex evolutionary models for which likelihoods are difficult orpractically impossible to compute (65, 66). The DIYABC program simulatescoalescent genealogies under user-specified evolutionary models usingparameters drawn from prior distributions and compares statistics sum-marizing various aspects of the simulated data with those of the observeddata. The similarity of summary statistics between observed and simulateddatasets is used to calculate posterior probabilities of competing evolu-tionary models and posterior distributions of parameters. In effect, modelsthat generate datasets with summary statistics close to the observed datahave higher probability. We used this method to evaluate posterior proba-bilities for alternate scenarios representing different possible evolutionaryrelationships among P. infestans populations in Mexico and the Andes. The useof ABC to evaluate alternate scenarios has been questioned owing to potentialproblems with using summary statistics, but DIYABC incorporates tests thataddress many of these concerns (67, 68). In particular, confidence in modelchoice can be evaluated empirically by calculating type I and II errors usingpseudo-observed datasets.

To narrow the number of evolutionary scenarios tested against one an-other, we used four sets of scenarios to inform a final fifth set of scenarios.Wegrouped the Andes isolates by clonal lineage to tease apart the evolutionaryhistory of each lineage. Isolates were assigned to a clonal lineage based ona previously determined RG-57 fingerprint (3) and confirmed with SSRgenotypes grouped using structure and k-means clustering (69). We firstsimulated the three different possible relationships among the Andeanlineages EC-1, PE-3, and PE-7 (SI Appendix, Fig. S9) using the settings inSI Appendix, Table S6A. We next examined 15 alternative relationships forthe US-1 lineage, the Toluca population, and each of the Andean lineages inturn (SI Appendix, Fig. S10 and Table S6B). Although the US-1 lineage hashad a global distribution, for these analyses we used only US-1 isolatescollected in the Andes, because our focus was on the genetic variation andevolution history of the Andes population. Based on the results of pre-liminary analyses, these scenarios included admixture events between line-ages as well as admixture with an unsampled population.

Finally, we used the scenarios with the highest posterior probabilities toconstruct three final scenarios (Fig. 4 and SI Appendix, Table S6C). This finalset tested admixture events generating the EC-1 and PE-3 lineages relativeto a scenario with no admixture. We removed the PE-7 lineage from thisfinal analysis, because there were too many plausible admixture eventsbehind the emergence of this lineage.

We tested prior distribution settings with small numbers of simulationsby running a principal components analysis on the summary statistics andcomparing the overlap of the simulated and observed data. Posteriorprobabilities and 95% confidence intervals were estimated by polychotomicweighted logistic regression on 1% of the simulations, as implemented inDIYABC. Model checking was conducted for the final three scenarios by es-timating type I and type II errors. In short, 500 pseudo-observed datasets (pods)were generated according to each evolutionary scenario using parametersdrawn from the sameprior distributions as used for the simulated genealogies.The posterior probability of each scenario was calculated for each pod. Type Ierror was estimated as the proportion of pods for which the correct scenariodid not receive the highest posterior probability. Type II errorwas calculated asthe proportion of pods erroneously assigned to a given scenario.

ACKNOWLEDGMENTS. We are grateful to the many colleagues who pro-vided isolates for this study. We thank Karan Fairchild, Kevin Myers, CarolinePress, and Naomi Williams for maintenance of the cultures and generaltechnical support. This research is supported in part by US Department ofAgriculture (USDA) Agricultural Research Service Grant 5358-22000-039-00D,USDA National Institute of Food and Agriculture Grant 2011-68004-30154(to N.J.G.), and the Scottish Government (D.E.L.C.).

Goss et al. PNAS | June 17, 2014 | vol. 111 | no. 24 | 8795

AGRICU

LTURA

LSC

IENCE

S

1. Yoshida K, et al. (2013) The rise and fall of the Phytophthora infestans lineage thattriggered the Irish potato famine. eLife 2:e00731.

2. Haverkort AJ, et al. (2008) Societal costs of late blight in potato and prospects ofdurable resistance through cisgenic modification. Potato Res 51:1871–4528.

3. Goodwin SB, Cohen BA, Fry WE (1994) Panglobal distribution of a single clonal line-age of the Irish potato famine fungus. Proc Natl Acad Sci USA 91(24):11591–11595.

4. Grünwald NJ, Flier WG (2005) The biology of Phytophthora infestans at its center oforigin. Annu Rev Phytopathol 43:171–190.

5. Gómez-Alpizar L, Carbone I, Ristaino JB (2007) An Andean origin of Phytophthorainfestans inferred from mitochondrial and nuclear gene genealogies. Proc Natl AcadSci USA 104(9):3306–3311.

6. Oliva RF, et al. (2010) Phytophthora andina sp nov., a newly identified heterothallicpathogen of solanaceous hosts in the Andean highlands. Plant Pathol 59(4):613–625.

7. Flier WG, et al. (2002) Phytophthora ipomoeae, a new homothallic species causinglate blight on Ipomoeae longipedunculata in the Toluca Valley of central Mexico.Mycol Res 106(7):848–856.

8. Goodwin SB, Legard DE, Smart CD, Levy M, Fry WE (1999) Gene flow analysis ofmolecular markers confirms that Phytophthora mirabilis and P. infestans are separatespecies. Mycologia 91(5):796–810.

9. Grünwald NJ, et al. (2001) Population structure of Phytophthora infestans in theToluca Valley region of Central Mexico. Phytopathology 91(9):882–890.

10. Flier WG, et al. (2003) The population structure of Phytophthora infestans from theToluca Valley in central Mexico suggests genetic differentiation between populationsfrom cultivated potato and wild Solanum species. Phytopathology 93(4):382–390.

11. Fry WE, et al. (1993) Historical and recent migrations of Phytophthora infestans:Chronology, pathways, and implications. Plant Dis 77(7):653–661.

12. Niederhauser JS (1991) Phytophthora infestans: The Mexican connection. Phytophthora,eds Lucas JA, Shattock RC, Shaw DS, Cooke LR (Cambridge Univ Press, Cambridge, UK), pp25–45.

13. Fry WE, et al. (1992) Population genetics and intercontinental migrations of Phy-tophthora infestans. Annu Rev Phytopathol 30:107–130.

14. Hijmans RJ, Spooner DM (2001) Geographic distribution of wild potato species. Am JBot 88(11):2101–2112.

15. Vleeshouwers VGAA, et al. (2011) Understanding and exploiting late blight resistancein the age of effectors. Annu Rev Phytopathol 49(1):507–531.

16. Goss EM, et al. (2011) The plant pathogen Phytophthora andina emerged via hy-bridization of an unknown Phytophthora species and the Irish potato famine path-ogen, P. infestans. PLoS ONE 6(9):e24543.

17. Adler NE, et al. (2004) Genetic diversity of Phytophthora infestans sensu lato inEcuador provides new insight into the origin of this important plant pathogen.Phytopathology 94(2):154–162.

18. Chacon MG, et al. (2006) Genetic structure of the population of Phytophthora in-festans attacking Solanum ochranthum in the highlands of Ecuador. Eur J PlantPathol 115(2):235–245.

19. Forbes GA, et al. (1997) Population genetic structure of Phytophthora infestans inEcuador. Phytopathology 87(4):375–380.

20. Grünwald NJ, Goss EM (2011) Evolution and population genetics of exotic and re-emerging pathogens: Novel tools and approaches. Annu Rev Phytopathol 49:249–267.

21. Tavare S (1984) Lines-of-descent and genealogical processes, and their applications inpopulation-genetics models. Adv Appl Prob 16(1):471–491.

22. Kingman JFC (1982) The coalescent. Stochastic Process Appl 13:235–248.23. Wakeley J (2009) Coalescent Theory: An Introduction (Roberts & Company, Greenwood

Village, CO).24. Lemey P, Rambaut A, Drummond AJ, Suchard MA (2009) Bayesian phylogeography

finds its roots. PLOS Comput Biol 5(9):e1000520.25. Hey J, Nielsen R (2004) Multilocus methods for estimating population sizes, migration

rates and divergence time, with applications to the divergence of Drosophila pseu-doobscura and D. persimilis. Genetics 167(2):747–760.

26. Miller N, et al. (2005) Multiple transatlantic introductions of the western corn root-worm. Science 310(5750):992.

27. Stukenbrock EH, Banke S, Javan-Nikkhah M, McDonald BA (2007) Origin and do-mestication of the fungal wheat pathogen Mycosphaerella graminicola via sympatricspeciation. Mol Biol Evol 24(2):398–411.

28. Lombaert E, et al. (2010) Bridgehead effect in the worldwide invasion of the bio-control harlequin ladybird. PLoS ONE 5(3):e9743.

29. Dutech C, et al. (2012) The chestnut blight fungus world tour: Successive introduction eventsfrom diverse origins in an invasive plant fungal pathogen. Mol Ecol 21(16):3931–3946.

30. Cooke DEL, Drenth A, Duncan JM, Wagels G, Brasier CM (2000) A molecular phy-logeny of Phytophthora and related oomycetes. Fungal Genet Biol 30(1):17–32.

31. Blair JE, Coffey MD, Park S-Y, Geiser DM, Kang S (2008) A multi-locus phylogeny forPhytophthora utilizing markers derived from complete genome sequences. FungalGenet Biol 45(3):266–277.

32. Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure usingmultilocus genotype data. Genetics 155(2):945–959.

33. Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian phylogenetics withBEAUti and the BEAST 1.7. Mol Biol Evol 29(8):1969–1973.

34. Blair JE, Coffey MD, Martin FN (2012) Species tree estimation for the late blightpathogen, Phytophthora infestans, and close relatives. PLoS ONE 7(5):e37003.

35. Cornuet JM, Ravigné V, Estoup A (2010) Inference on population history and modelchecking using DNA sequence and microsatellite data with the software DIYABC(v1.0). BMC Bioinformatics 11:401.

36. Guillemaud T, Beaumont MA, Ciosi M, Cornuet J-M, Estoup A (2010) Inferring in-troduction routes of invasive species using approximate Bayesian computation onmicrosatellite data. Heredity (Edinb) 104(1):88–99.

37. Gallegly ME, Galindo J (1958) Mating types and oospores of Phytophthora infestans innature in Mexico. Phytopathology 48:274–277.

38. Flier WG, Grünwald NJ, Fry WE, Turkensteen LJ (2001) Formation, production andviability of oospores of Phytophthora infestans isolates from potato and Solanumdemissum in the Toluca Valley, central Mexico. Mycol Res 105(8):998–1006.

39. Cooke DEL, et al. (2012) Genome analyses of an aggressive and invasive lineage of theIrish potato famine pathogen. PLoS Pathog 8(10):e1002940.

40. Spielman LJ, et al. (1991) A second world-wide migration and population displace-ment of Phytophthora infestans? Plant Pathol 40(3):422–430.

41. Li Y, et al. (2012) Population structure of Phytophthora infestans in China: Geographicclusters and presence of the EU genotype Blue_13. Plant Pathol 62:932–942.

42. Chowdappa P, et al. (2012) Emergence of 13_A2 blue lineage of Phytophthora in-festans was responsible for severe outbreaks of late blight on tomato in south-westIndia. J Phytopathol 161:49–58.

43. Goodwin SB, et al. (1998) Genetic change within populations of Phytophthora in-festans in the United States and Canada during 1994 to 1996: Role of migration andrecombination. Phytopathology 88(9):939–949.

44. Tooley PW, Therrien CD, Ritch DL (1989) Mating type, race composition, nuclear DNAcontent, and isozyme analysis of Peruvian isolates of Phytophthora infestans. Phy-topathology 79:478–481.

45. Stukenbrock EH, McDonald BA (2008) The origins of plant pathogens in agro-ecosystems.Annu Rev Phytopathol 46:75–100.

46. Zaffarano PL, McDonald BA, Linde CC (2008) Rapid speciation following recent hostshifts in the plant pathogenic fungus Rhynchosporium. Evolution 62(6):1418–1436.

47. Couch BC, et al. (2005) Origins of host-specific populations of the blast pathogenMagnaporthe oryzae in crop domestication with subsequent expansion of pandemicclones on rice and weeds of rice. Genetics 170(2):613–630.

48. Goodwin SB, Sujkowski LS, Fry WE (1995) Rapid evolution of pathogenicity withinclonal lineages of the potato late blight disease fungus. Phytopathology 85:669–676.

49. McDonald BA, Linde C (2002) Pathogen population genetics, evolutionary potential,and durable resistance. Annu Rev Phytopathol 40:349–379.

50. Lees AK, et al. (2006) Novel microsatellite markers for the analysis of Phytophthorainfestans populations. Plant Pathol 55:311–319.

51. Stephens M, Smith NJ, Donnelly P (2001) A new statistical method for haplotype re-construction from population data. Am J Hum Genet 68(4):978–989.

52. Szpiech ZA, Jakobsson M, Rosenberg NA (2008) ADZE: Allelic Diversity Analyzer,version 1.0. http://rosenberglab.bioinformatics.med.umich.edu/adze.html.

53. Falush D, Stephens M, Pritchard JK (2003) Inference of population structure usingmultilocus genotype data: Linked loci and correlated allele frequencies. Genetics164(4):1567–1587.

54. Earl DA, Vonholdt BM (2012) STRUCTURE HARVESTER: A website and program forvisualizing STRUCTURE output and implementing the Evanno method. Conserv GenetResour 4(2):359–361.

55. Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individualsusing the software STRUCTURE: A simulation study. Mol Ecol 14(8):2611–2620.

56. Brown AHD, Feldman MW, Nevo E (1980) Multilocus structure of natural populationsof Hordeum spontaneum. Genetics 96(2):523–536.

57. Agapow P-M, Burt A (2001) Indices of multilocus linkage disequilibrium. Mol EcolNotes 1(1-2):101–102.

58. Kamvar ZN, Tabima JF, Grünwald NJ (2014) Poppr: An R package for genetic analysisof populations with clonal, partially clonal, and/or sexual reproduction. PeerJ 2:e281.

59. R Development Core Team (2013) R: A Language and Environment for StatisticalComputing (R Foundation for Statistical Computing, Vienna, Austria).

60. Drummond AJ, Rambaut A (2007) BEAST: Bayesian evolutionary analysis by samplingtrees. BMC Evol Biol 7:214.

61. Darriba D, Taboada GL, Doallo R, Posada D (2012) jModelTest 2: More models, newheuristics and parallel computing. Nat Methods 9(8):772.

62. Keane TM, Creevey CJ, Pentony MM, Naughton TJ, Mclnerney JO (2006) Assessmentof methods for amino acid matrix selection and their use on empirical data shows thatad hoc assumptions for choice of matrix are not justified. BMC Evol Biol 6:29.

63. WickhamH (2009) ggplot2: Elegant Graphics for Data Analysis (Springer, New York), p 212.64. Hey J (2010) Isolation with migration models for more than two populations.Mol Biol

Evol 27(4):905–920.65. Beaumont MA (2010) Approximate Bayesian computation in evolution and ecology.

Annu Rev Ecol Evol Syst 41(1):379–406.66. Csilléry K, Blum MGB, Gaggiotti OE, François O (2010) Approximate Bayesian Com-

putation (ABC) in practice. Trends Ecol Evol 25(7):410–418.67. Robert CP, Cornuet JM, Marin JM, Pillai NS (2011) Lack of confidence in approximate

Bayesian computation model choice. Proc Natl Acad Sci USA 108(37):15112–15117.68. Bertorelle G, Benazzo A, Mona S (2010) ABC as a flexible framework to estimate

demography over space and time: Some cons, many pros. Mol Ecol 19(13):2609–2625.69. Meirmans P, Van Tienderen P (2004) Genotype and Genodive: Two programs for the

analysis of genetic diversity of asexual organisms. Mol Ecol Notes 4:792–794.

8796 | www.pnas.org/cgi/doi/10.1073/pnas.1401884111 Goss et al.

! 1!

Supporting Information Supporting*Methods*............................................................................................................................*1!Supporting*Results*..............................................................................................................................*2!References*Cited*...................................................................................................................................*3!Supporting*Tables*................................................................................................................................*4!Supporting*Figures*............................................................................................................................*19!

Supporting Methods Genetree analysis. We revisited the coalescent analysis of Gómez-Alpizar et al. (1) using their data and ours to investigate the sensitivity of root inference to estimates of migration rates. Our concern was that there is little power to estimate migration rates with only a single locus, yet these rates were used to infer the rooting of structured coalescent trees in the program Genetree (2). Furthermore, Gómez-Alpizar et al. considered migration to and from South America, which confounds migration from Mexico and other global populations. P. infestans is transported transcontinentally and intercontinentally via infected seed tubers, and there is a vibrant commercial seed tuber trade from Northwest Europe with known pathways of migration from Europe to South America.

We revised the Genetree analysis of the P3 and P4 mtDNA regions using sequences from Gómez-Alpizar et al. (1) for isolates from Mexico and the Andes but removing the haplotypes representing the P. andina isolates (haplotype 1c). The topology and rooting of the coalescent tree for the P. infestans haplotypes remained the same. We inferred the maximum likelihood value of theta and a symmetric migration rate for the revised data set. We then used 1×107 simulations to examine the inferred subpopulation of the most common recent ancestor (MRCA) of the sample while varying theta and symmetric migration rates using three sets of runs, each set using a different starting seed. For the RAS nuclear gene, we repeated the Gómez-Alpizar et al. (1) analysis while varying migration rates. We used their data set and value of theta. We conducted two runs of 1×107 simulations for each set of migration rates. We varied migration rates from low and symmetric to the values used by Gómez-Alpizar et al. We also examined the coalescent history of the RAS locus using our sequences from Mexico and the Andes. The tree was rooted using ancestral states obtained from P. ipomoeae, P. mirabilis, and P. phaseoli. We used four runs of 1×106 or 1×107 simulations and three symmetric migration rates using two different values of theta (1.5 and 2.0) and four starting seeds.

Migration scenarios from SSR genotypes. The program Migrate version 3.3 was used to examine recent migration between the Andes and Mexico, using SSR genotypes. Three migration models were compared (3): bidirectional migration, migration from Mexico to the Andes only, and migration from the Andes to Mexico only. The runs used Brownian motion approximation for the stepwise mutation model, initial parameter values from FST, and mutation rates per locus estimated from the data. Bayesian inference across 10 replicates used slice sampling, uniform prior distributions from 0 to 1000 for both theta and M parameters, and four chains (temperatures: 1,000,000, 3.0, 1.5, 1.0) in which the cold chain was sampled at 10,000

! 2!

steps at 200 step increments after a burnin of 500,000 steps. Models were evaluated using Bezier approximation of log marginal maximum likelihoods.

Supporting Results

Dependence of Genetree results on migration rates. When we removed P. andina from the Gómez-Alpizar et al. mitochondrial data set and included only the Mexican and South American isolates, the root location was uncertain such that there was equal probability of rooting in Mexico or South America when migration rates were symmetric (Table S7A). Asymmetric migration rates increased the probability of the root in the population that was the source of migration to around 0.95, irrespective of which population was assigned the higher emigration rate.

For the RAS locus, the location of the coalescent root was dependent on migration rates as well (Table S7B). When we used the Gómez-Alpizar et al. data under low symmetric migration rates the rooting was uncertain, and asymmetric migration rates affected the inferred population of origin. The migration rates used by Gómez-Alpizar et al. produced a high probability for a South American root, which replicates their result. The opposite condition produced a similarly high probability for a non-South American root. Therefore, both mitochondrial and RAS analyses were highly dependent on choice of migration parameters.

We used our data set to examine the root location of RAS for Mexican and Andean isolates only. Using RAS sequences from P. andina, P. mirabilis, P. ipomoeae, and P. phaseoli, we were able to unambiguously assign ancestral states to each segregating site and these ancestral states were used to root the P. infestans RAS tree. The coalescent history of the gene was simulated on the rooted topology to infer the location of the root in Mexico or the Andes. Moderate and inconclusive probabilities for root location were observed under nearly all conditions examined, including three symmetric migration rates and two values of theta (Table S8).

Key to the Gómez-Alpizar et al. analysis was their finding of a higher migration rate from South America to non-South American populations (i.e. Mexico, USA, and Ireland) than vice-versa using the nuclear RAS locus. Given this result, we used our SSR data to test among models of migration between the Andes and Mexico using the program Migrate (3). Multiple loci are expected to provide more robust estimates of migration rates for the nuclear genome than can be obtained from a single locus. This analysis revealed a higher migration rate from Mexico to the Andes than from the Andes to Mexico (M = 11.7 vs. 0.3, respectively). Bayesian comparison of migration models strongly supported unidirectional migration from Mexico to the Andes (Bezier approximation of the log marginal likelihood = -11,055) compared to the opposite scenario of migration from the Andes to Mexico (-12,243) or bidirectional migration (-52,357). This result is consistent with the known directions of trade in commercial seed tubers. !

! 3!

References Cited 1. Gómez-Alpizar L, Carbone I, & Ristaino JB (2007) An Andean origin of Phytophthora

infestans inferred from mitochondrial and nuclear gene genealogies. Proc Natl Acad Sci U S A 104(9):3306-3311.

2. Griffiths RC & Tavaré S (1994) Ancestral inference in population genetics. Statistical science 9(3):307-319.

3. Beerli P & Palczewski M (2010) Unified framework to evaluate panmixia and migration direction among multiple sampling locations. Genetics 185(1):313-U463.

4. Nei M (1987) Molecular Evolutionary Genetics (Columbia University Press, New York) p 512

5. Watterson GA (1975) On the number of segregating sites in genetic models without recombination. Theoretical Population Biology 7:256-276.

6. Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123(3):585-595.

7. Fu Y-X & Li W-H (1993) Statistical tests of neutrality of mutations. Genetics 133(3):693-709.

! 4!

Supporting Tables Table S1. Isolates used in the study.

ID Original name Year sampled Country State Host Source1

Lineage

P. infestans

PiCO01 1011 Colombia Cundinamarca S. tuberosum Restrepo EC-1 PiCO02 1063 Colombia Cundinamarca S. tuberosum Restrepo EC-1 PiCO03 1064 Colombia Cundinamarca S. tuberosum Restrepo EC-1

PiCO04 1068 Colombia Cundinamarca S. tuberosum Restrepo EC-1

PiCO05 4084 Colombia Cundinamarca Physalis peruviana Restrepo US-8

PiEC01 EC_3843 2004 Ecuador potato Forbes EC-1

PiEC02 EC_3527 2002 Ecuador S. andreanum Forbes EC-1

PiEC03 EC_3626 2003 Ecuador potato Forbes EC-1

PiEC06 EC_3841 2004 Ecuador S. habrochaites Forbes US-1

PiEC07 EC_3921 2006 Ecuador S. jugandifolium Forbes US-1

PiEC08 EC_3774 2004 Ecuador S. ochanthum Forbes US-1

PiEC10 EC_3378 2001 Ecuador S. lycopersicum Forbes US-1

PiEC11 EC_3381 2001 Ecuador S. lycopersicum Forbes US-1

PiEC12 EC_3150 1997 Ecuador S. muricatum Forbes US-1

PiEC13 EC_3520 2002 Ecuador S. muricatum Forbes US-1

PiEC14 EC_3809 2004 Ecuador S. caripense Forbes US-1

PiES01 2005_10 2005 Estonia potato Fry

PiES02 2005_17 2005 Estonia potato Fry

PiES03 2005_19 2004 Estonia potato Fry

PiES04 2004_4 2004 Estonia potato Fry

PiES05 2004_16 2004 Estonia potato Fry

PiHU02 Josze_S32 Hungary potato Bakonyi

PiMX01 MX010006 Mexico potato Fry

PiMX02 MX010046 Mexico potato Fry

! 5!

PiMX03 MX980211 1998 Mexico potato Fry

PiMX04 MX980230 1998 Mexico potato Fry

PiMX05 MX980317 1998 Mexico potato Fry

PiMX06 MX980352 1998 Mexico potato Fry

PiMX07 MX980400 1998 Mexico potato Fry

PiMX10 PIC97008 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX11 PIC97066 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX12 PIC97106 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX13 PIC97111 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX14 PIC97130 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX15 PIC97136 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX16 PIC97146 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX17 PIC97149 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX18 PIC97153 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX19 PIC97159 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX20 PIC97187 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX21 PIC97310 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX22 PIC97318 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX23 PIC97335 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX24 PIC97340 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX25 PIC97389 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX26 PIC97392 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX27 PIC97423 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX28 PIC97432 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX29 PIC97438 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX30 PIC97442 1997 Mexico Toluca potato Flier/Grünwald/PRI

PiMX40 PIC97716 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX41 PIC97724 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX42 PIC97727 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX43 PIC97744 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX44 PIC97748 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

! 6!

PiMX45 PIC97749 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX46 PIC97750 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX47 PIC97751 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX48 PIC97785 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX49 PIC97791 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX50 PIC97793 1997 Mexico Toluca S. demissum Flier/Grünwald/PRI

PiMX61 Tlax 701 2007 Mexico Tlaxcala potato Fernández Pavia

PiMX62 Tlax 715 2007 Mexico Tlaxcala potato Fernández Pavia

PiMX63 Tlax 722 2007 Mexico Tlaxcala potato Fernández Pavia

PiMX64 Tlax 728 2007 Mexico Tlaxcala potato Fernández Pavia

PiMX65 Tlax 740 2007 Mexico Tlaxcala potato Fernández Pavia

PiMX66 Tlax 748 2007 Mexico Tlaxcala potato Fernández Pavia

PiMX67 Tlax 756 2007 Mexico Tlaxcala potato Fernández Pavia

PiMX71 T48 2003 Mexico Tlaxcala potato Fernández Pavia

PiMX72 T68 2003 Mexico Tlaxcala potato Fernández Pavia

PiNL01 NL_01096 2001 Netherlands potato Kessel

PiNL02 NL_96259 1996 Netherlands potato Kessel

PiPE01 BTLM 004 1997 Peru Lima NA Forbes PE-7

PiPE02 PHU 076 2003 Peru Huánuco potato Forbes EC-1

PiPE03 PHU 079 2003 Peru Huánuco potato Forbes EC-1

PiPE04 PPI 015 2000 Peru Piura S. huancabambense Forbes EC-1

PiPE05 PTS 031 1998

Peru Cajamarca S. caripense Forbes EC-1

PiPE06 1696 1995 Peru Arequipa potato Forbes PE-3

PiPE07 PCA 004 1999 Peru Cajamarca potato Forbes PE-3

PiPE08 PCZ 024 1997 Peru Cuzco potato Forbes EC-1

PiPE09 PCZ 080 1997 Peru Cuzco potato Forbes EC-1

PiPE10 PSR 001 2005 Peru Junín NA Forbes EC-1

PiPE11 PCA 020 1999 Peru Cajamarca NA Forbes PE-7

PiPE12 PLI 003 1999 Peru Lima NA Forbes PE-7

PiPE13 PLI 036 2000 Peru Lima S. wittmackii Forbes PE-7

! 7!

PiPE14 POX 100 2003 Peru Pasco potato Forbes PE-7

PiPE20 PCA 025 1999 Peru Cajamarca NA Forbes US-1

PiPE21 PPI 009 2000 Peru Piura S. caripense Forbes US-1

PiPE22 PPI 013 2000 Peru Piura S. caripense Forbes US-1

PiPE23 PPI 014 2000 Peru Piura S. caripense Forbes US-1

PiPE24 PPI 023 2000 Peru Piura S. caripense Forbes US-1

PiPE25 PPI 028 2000 Peru Piura S. caripense Forbes US-1

PiPE26 PPU 048 1997 Peru Puno potato Forbes PE-3

PiPE27 PPU 097 1997 Peru Puno potato Forbes US-1

PiPE28 PCA 006 1999 Peru Cajamarca potato Forbes PE-3

PiPE29 PCA 010 1999 Peru Cajamarca potato Forbes PE-3

PiPO01 MP_618 2005 Poland potato Lebecka

PiPO02 MP_622 2005 Poland potato Lebecka

PiSA01 SA960008 1996 South Africa potato Fry US-1

PiSW01 SE_03058 2003 Sweden potato Andersson

PiSW02 SE_03087 2003 Sweden potato Andersson

PiUK01 2006_3984C 2006 UK potato Cooke EU_1_A1

PiUK02 2006_4012F 2006 UK potato Cooke EU_3_A2

PiUK03 2006_3928A 2006 UK potato Cooke EU_13_A2

PiUK04 2006_4132B 2006 UK potato Cooke EU_13_A2

PiUK05 2006_3888A 2006 UK potato Cooke EU_2_A1

PiUK06 2007_5866B 2007 UK potato Cooke EU_5_A1

PiUK07 2006_4388D 2006 UK potato Cooke EU_17_A2

PiUK08 2006_4100A 2006 UK potato Cooke EU_6_A1

PiUK09 2006_4440C 2006 UK potato Cooke EU_10_A2

PiUK10 2006_4232E 2006 UK potato Cooke EU_8_A1

PiUS08 US040009 2004 USA potato Fry US-8

PiUS11 US050007 2005 USA tomato Fry US-11

PiUS12 US940494 1994 USA tomato Fry US-12

PiUS17 US970001 1997 USA tomato Fry US-17

PiVT01 Vn02-076 2002 Vietnam tomato Le US-1

! 8!

PiVT02 Vn02-106 2002 Vietnam tomato Le US-1

PiVT03 Vn03-416 2003 Vietnam potato Le US-1

PiVT04 Vn03-590 2003 Vietnam potato Le US-1

P. mirabilis

PmMX01 CBS 136.86 Mexico NA CBS

PmMX02 CBS 678.85 Mexico NA CBS

PmMX03 DF 409 Mexico NA Fernández Pavia

PmMX04 G 11-3 1998 Mexico NA Flier/Grünwald/PRI

PmMX05 00M 410 2000 Mexico NA Fernández Pavia

PmMX06 P 3001 Mexico NA Flier/Grünwald/PRI

PmMX07 P 3006 Mexico NA Flier/Grünwald/PRI

PmMX08 P mirabilis DF 07 2007 Mexico NA Fernández Pavia

PmMX09 P. mirabilis Mich 2003 Mexico NA Fernández Pavia

PmMX10 WF014/PIC99114 1999 Mexico NA Flier/Grünwald/PRI

PmMX11 WF035/PIC99135 1999 Mexico NA Flier/Grünwald/PRI

P. ipomoeae

PoMX01 00Ip5 2000 Mexico NA Fernández Pavia

PoMX02 Ipom1-2 1999 Mexico NA Flier/Grünwald/PRI

PoMX03 Ipom2-1 1999 Mexico NA Flier/Grünwald/PRI

PoMX04 Ipom2-4 1999 Mexico NA Flier/Grünwald/PRI

PoMX05 Ipom3-3 1999 Mexico NA Flier/Grünwald/PRI

PoMX06 Ipom6 1999 Mexico NA Flier/Grünwald/PRI

PoMX07 P. ipomoeae 1999 Mexico NA Fernández Pavia

P. andina

PaEC01 EC_3189 1998 Ecuador Anarrichomenum Forbes

PaEC02 EC_3399 2001 Ecuador Anarrichomenum Cooke

PaEC03 EC_3818 2004 Ecuador Anarrichomenum Cooke

PaEC04 EC_3821 2004 Ecuador Anarrichomenum Forbes

PaEC05 EC_3780 2004 Ecuador S. hispidum Forbes

PaEC06 EC_3655 2003 Ecuador S. hispidum Forbes

! 9!

PaEC07 EC_3163 1998 Ecuador Anarrichomenum Forbes

PaEC08 EC_3510 2002 Ecuador S. betaceum Forbes

PaEC09 EC_3540 2002 Ecuador Anarrichomenum Forbes

PaEC10 EC_3561 2002 Ecuador S. quitoense Forbes

PaEC11 EC_3563 2002 Ecuador S. quitoense Forbes

PaEC12 EC_3678 2003 Ecuador Anarrichomenum Forbes

PaEC13 EC_3836 2004 Ecuador S. betaceum Forbes

PaEC14 EC_3860 2005 Ecuador Torva Forbes

PaEC15 EC_3864 2005 Ecuador Torva Forbes

PaEC16 EC_3865 2005 Ecuador S. jugandifolium Forbes

PaEC17 EC_3936 2006 Ecuador S. ochanthum Forbes

PaPE01 POX 102 2003 Peru S. betaceum Forbes

PaPE02 POX 103 2003 Peru S. betaceum Forbes

P. phaseoli

PpUS01 CBS 556.88 CBS

PpUS02 P10150 USA Delaware Coffey 1 Isolates were contributed by the authors and the following colleagues: Sampled by W. G. Flier and N. J. Grünwald and curated by Geert Kessel, Plant Research International (PRI), Netherlands; Geert Kessel, Plant Research International (PRI), Netherlands; Silvia Fernández Pavia, Universidad Michoacana de San Nicolás de Hidalgo, Mexico; Michael Coffey, University of California Riverside, USA; Vihn Hong Le and Arne Hermansen, Norwegian Institute for Agricultural and Environmental Research, Norway; Björn Andersson, Swedish University of Agricultural Sciences, Sweden; Renata Lebecka Młochow Research Centre, Poland; Jozsef Bakonyi, Academy of Agricultural Sciences, Hungary.

  10  

Table S2. Molecular clock and mutation rates obtained in this study for each locus for (A) Clade 1c and (B) P. infestans datasets. See Table S3 for indices of nucleotide variation at each locus.

A. Clade 1c Locus Clock Clock Rate UCLD stdev β-tubulin Strict 0.824 0.04 RAS Log Normal 1.839 2.2 Trp1 Strict 1.73 0.03 PITG_11126 Strict 4.09 0.02 B. P. infestans Locus Clock Clock Rate UCLD stdev β-tubulin Strict 0.54 0.02 RAS Log Normal 2.25 1.94 Trp1 Strict 1.73 0.012 PITG_11126 Strict 3.03 0.01

! 11!

Table S3. Nucleotide variation by locus, species, and sampling location. Statistics given are number of individuals (Nind), number of sequences (Nseq), length of alignment excluding gaps (L), segregating sites (S), number of heterozygous sites (Het), number of haplotypes (Hap), average pairwise nucleotide diversity (π), Watterson’s theta (θw), Tajima’s D, and Fu and Li’s D* and F* (4-7). Locus Species/Popn Nind Nseq L S Het Hap π θW Tajima’s

D Fu & Li’s D*

Fu & Li’s F*

PITG_11126 P. infestans 119 242 764 13 13# 10 0.00286 2.144 0.04 0.79 0.61 Europe 22 46 764 7 7# 5 0.00294 1.593 1.10 1.25 1.41 Mexico 48 96 764 9 9# 6 0.00154 1.752 -0.82 -0.17 -0.47 S. America 40 82 764 11 11# 8 0.00394 2.210 0.97 0.79 1.01 United States 4 8 779 2 2 2 0.00138 0.771 1.45 1.11 1.30 Africa/Asia 5 10 779 3 3 2 0.00214 1.060 2.06* 1.15 1.53 P. andina 19 38 748 16 16# 3 0.01040 3.808 3.39*** 1.58** 2.56** P. ipomoeae 7 14 787 2 1 3 0.00052 0.629 -0.96 -0.45 -0.66 P. mirabilis 11 22 778 11 11 3 0.00593 4.610 1.83 1.44* 1.80** P. phaseoli 2 4 779 0 0 1 0 0 NA NA NA β-tubulin P. infestans 118 238 1590 9 9# 8 0.00138 1.488 1.06 -0.43 0.13 Europe 22 44 1590 6 6# 3 0.00061 1.379 -0.79 1.18 0.67 Mexico 48 96 1590 7 7# 3 0.00185 1.363 2.75** 1.20 2.03** S. America 39 80 1590 9 9# 7 0.00113 1.817 -0.03 -0.84 -0.67 United States 4 8 1590 5 5# 3 0.00124 1.928 0.08 0.75 0.65 Africa/Asia 5 10 1590 0 0 1 0 0 NA NA NA P. andina 19 38 1590 23 23 2 0.00743 5.474 3.92*** 1.70** 2.89** P. ipomoeae 7 14 1590 2 2 2 0.00018 0.629 -1.48 -1.83 -1.97 P. mirabilis 11 22 1590 16 15# 6 0.00469 4.389 2.54** 1.54** 2.14** P. phaseoli 2 4 1590 0 0 1 0 0 NA NA NA Trp1 P. infestans 117 234 813 9 7 9 0.00183 1.492 -0.004 0.42 0.32 Europe 22 44 813 4 3 4 0.00185 0.920 1.48 -0.06 0.47 Mexico 48 96 813 6 4 5 0.00144 1.168 -0.20 1.12 0.81 S. America 38 76 813 6 6 6 0.00211 1.224 0.95 0.22 0.54 United States 4 8 813 4 4 4 0.00180 1.543 -0.22 -0.18 -0.21 Africa/Asia 5 10 813 3 3 2 0.00205 1.060 2.06* 1.15 1.53 P. andina 19 38 813 10 10 3 0.00293 2.380 3.48*** 1.40* 2.42** P. ipomoeae 7 14 812 0 0 1 0 0 NA NA NA P. mirabilis 11 22 813 1 1 2 0.00011 0.274 -1.16 -1.57 -1.68 P. phaseoli 2 4 813 0 0 1 0 0 NA NA NA RAS P. infestans 117 234 762 15 15# 13 0.00364 2.487 0.290 0.311 0.364 Europe 21 42 766 10 10 6 0.00569 2.324 2.585* 1.40 2.096** Mexico 47 94 766 11 11# 8 0.00164 2.150 -1.09 -2.51* -2.39* S. America 40 80 762 10 10 5 0.00377 2.019 1.12 1.38 1.53 United States 4 8 766 11 11 4 0.00643 4.242 0.808 1.52** 1.50 Africa/Asia 5 10 766 1 1 2 0.00205 1.060 2.06* 1.15 1.53 P. andina 17 34 765 23 23# 4 0.01433 5.625 3.28*** 1.39 2.37** P. ipomoeae 7 14 768 3 0 2 0.00103 0.943 -0.49 1.07 0.76 P. mirabilis 10 20 766 6 6# 3 0.00091 1.973 -2.13* -3.08** -3.25** P. phaseoli 1 2 764 0 0 1 0 0 NA NA NA

! 12!

#Also heterozygous for one or more indel. * P < 0.05 ** P < 0.02 *** P < 0.001

! 13!

Table S4. Results from testing relationships among Toluca population, US1 and Andean lineages as illustrated in Figure S10. Andean lineage Scenario Admixed population Probability 95% CI PE3 1 - 0.0040 [0.0028,0.0052] 2 - 0.0071 [0.0058,0.0084] 3 - 0.0066 [0.0055,0.0077] 4 PE3 0.0014 [0.0010,0.0017] 5 Toluca 0.0062 [0.0051,0.0073] 6 US1 0.0009 [0.0008,0.0011] 7 PE3 0.6709 [0.6478,0.6939] 8 PE3 0.1094 [0.0994,0.1194] 9 US1 0.0023 [0.0011,0.0034] 10 US1 0.0052 [0.0034,0.0070] 11 Toluca 0.0020 [0.0009,0.0032] 12 Toluca 0.1302 [0.1093,0.1511] 13 PE3 0.0474 [0.0410,0.0537] 14 US1 0.0011 [0.0004,0.0017] 15 Toluca 0.0053 [0.0034,0.0073] PE7 1 - 0.0089 [0.0055,0.0124] 2 - 0.0327 [0.0265,0.0389] 3 - 0.0141 [0.0107,0.0176] 4 PE7 0.0038 [0.0025,0.0051] 5 Toluca 0.0093 [0.0073,0.0112] 6 US1 0.0191 [0.0148,0.0234] 7 PE7 0.3656 [0.3393,0.3919] 8 PE7 0.2893 [0.2656,0.3131] 9 US1 0.0016 [0.0003,0.0030] 10 US1 0.0248 [0.0145,0.0350] 11 Toluca 0.0121 [0.0032,0.0211] 12 Toluca 0.0880 [0.0673,0.1087] 13 PE7 0.1242 [0.1069,0.1416] 14 US1 0.0012 [0.0003,0.0022] 15 Toluca 0.0050 [0.0021,0.0079] EC1 1 - 0.2524 [0.2335,0.2713] 2 - 0.0293 [0.0260,0.0326] 3 - 0.0575 [0.0519,0.0631] 4 EC1 0.0908 [0.0830,0.0987] 5 Toluca 0.0843 [0.0773,0.0913] 6 US1 0.0216 [0.0192,0.0239] 7 EC1 0.0320 [0.0276,0.0364] 8 EC1 0.3060 [0.2811,0.3309] 9 US1 0.0071 [0.0053,0.0088] 10 US1 0.0234 [0.0188,0.0280] 11 Toluca 0.0177 [0.0144,0.0210] 12 Toluca 0.0140 [0.0116,0.0163] 13 EC1 0.0088 [0.0069,0.0106] 14 US1 0.0517 [0.0420,0.0614] 15 Toluca 0.0035 [0.0026,0.0043]

! 14!

Table S5. Nucleotide substitution model for each locus as inferred from jModelTest and ModelGenerator. The appropriate most similar model available in BEAST was used for each analysis.

A. Clade 1c

Locus jModelTest ModelGenerator Model Used β-tubulin HKY HKY HKY RAS JC JC JC Trp1 K80 HKY HKY PITG_11126 HKY HKY HKY B. P. infestans Locus jModelTest ModelGenerator Model Used β-tubulin HKY HKY HKY RAS F81 JC JC Trp1 HKY HKY HKY PITG_11126 HKY HKY HKY

! 15!

Table S6. Settings used for DIYABC analyses. All runs used the same summary statistics: within population statistics were number of segregating sites, mean of pairwise differences, variance of pairwise differences, and number of private segregating sites; between population statistics were number of segregating sites, mean of pairwise differences, and FST. Posterior probabilities of scenarios were calculated using the closest 1% of simulated datasets using logistic regression. A. Prior distributions for scenarios in Figure S9. Parameter Shape Min.–Max. Increment Population size EC-1 Uniform 1–200000 1 PE-3 Uniform 1–200000 1 PE-7 Uniform 1–200000 1 Ancestor of two lineages Uniform 1–500000 1 Ancestor of three lineages Uniform 1–1000000 1 Time since divergence1 t1: two lineages Uniform 1–300000 1 t2: all three lineages Uniform 1–500000 1 Nucleotide sequence evolution2 Mean mutation rate Uniform 1.00×10-10 –

1.00×10-8

Gamma distribution Uniform 1.00×10-11 – 1.00×10-7

2.00

1 An additional condition was put on the prior distributions such that t2>t1. 2 The Kimura 2-parameter nucleotide substitution model was used with invariant sites=90 and gamma rate variation parameter α=0.050. B. Prior distributions for scenarios in Figure S10. Parameter Shape Min.–Max. Increment Population size Toluca Uniform 1–1000000 1 EC-1 Uniform 1–300000 1 PE-3 Uniform 1–200000 1 PE-7 Uniform 1–200000 1 US-1 Uniform 1–500000 1 Unsampled population Uniform 1–2000000 1 Ancestor of two lineages Uniform 1–2000000 1 Time since divergence3 t1: two populations (scenarios 1-3) Uniform 1–200000 1 t2: two or three populations Uniform 1–500000 1 t3: sampled & unsampled populations

(scenarios 7-15) Uniform 1–10000000 1

Admixture events Proportion from parent population Uniform 0.001–0.999 0.001 ta1: timing since admixture event

(scenarios 4-12) Uniform 1–300000 1

ta2: timing since admixture event (scenarios 13-15)

Uniform 1–1000000 1

! 16!

Nucleotide sequence evolution4 Mean mutation rate Uniform 1.00×10-10 –

1.00×10-8

Gamma distribution Uniform 1.00×10-11 – 1.00×10-7

2.00

3 Additional conditions were put on the prior distributions to force the timing of events into the tree structure shown in Figure S9, including placing admixture events before or after population splits and ordering of the population splits such that t2>t1, t2>ta1, t3>ta1, ta2>t2, and t3>ta2. 4 The Kimura 2-parameter nucleotide substitution model was used with invariant sites=90 and gamma rate variation parameter α=0.050. C. Prior distributions for scenarios in Figure 5. Parameter Shape Min.–Max. Increment Population size Toluca Uniform 1–1000000 1 EC-1 Uniform 1–400000 1 PE-3 Uniform 1–200000 1 US-1 Uniform 1–400000 1 Unsampled population Uniform 1–1000000 1 Ancestor to Toluca & EC-1 Uniform 10–1000000 1 Ancestor to Toluca & US-1 Uniform 10–1000000 1 Time since divergence5 t1: Toluca–EC-1 Uniform 1–200000 1 t2: Toluca–US-1 Uniform 1–600000 1 t3: All populations Uniform 1–1000000 1 Admixture events Proportion from parent population Uniform 0.001–0.999 0.001 Timing since admixture event Uniform 1–200000 1 Nucleotide sequence evolution6 Mean mutation rate Uniform 1.00×10-10 –

1.00×10-8

Gamma distribution Uniform 1.00×10-11 – 1.00×10-7

2.00

5 Additional conditions were put on the prior distributions to force the timing of events into the tree structure shown in Figure 5, including placing admixture events before or after population splits and ordering of the population splits such that t3>t1, t3>t2. 6 The Kimura 2-parameter nucleotide substitution model was used with invariant sites=90 and gamma rate variation parameter α=0.050.

! 17!

Table S7. Inferred root subpopulation using three starting seeds for (A) the mitochondrial P3 and P4 regions using only isolates from Mexico and the Andes, and (B) the nuclear RAS locus, using the data from Gómez-Alpizar et al. (1). A. P3 and P4 Seed 1 Seed 2 Seed 3 Theta Migration Root

location Prob. Root

location Prob.

Root location

Prob.

1.8 0.3, 0.3 Mexico 0.505 Mexico 0.500 Andes 0.501 1.0, 1.0 Andes 0.501 Andes 0.505 Mexico 0.508 5.0, 5.0 Mexico 0.559 Mexico 0.556 Mexico 0.510 10.0, 10.0 Mexico 0.587 Mexico 0.577 Andes 0.669 2.0 1.0, 1.0 Andes 0.504 Andes 0.501 Mexico 0.503 5.0, 5.0 Andes 0.513 Andes 0.531 Andes 0.562 10.0, 10.0 Andes 0.689 Mexico 0.727 Andes 0.549 B. RAS Seed 1 Seed 2 Seed 3 Theta Migration Root

location Prob. Root

location Prob.

Root location

Prob.

2.0 0.5, 0.5 SA 0.515 NSA 0.537 SA 0.535 5.0, 5.0 SA 0.514 SA 0.582 SA 0.847 10.0, 10.0 NSA 0.889 NSA 0.875 SA 0.686 5.0, 10.0 SA 0.916 SA 0.949 SA 0.985 10.0, 5.0 NSA 0.679 NSA 0.902 SA 0.935 4.3, 14.6 * SA 0.965 SA 0.992 SA 0.970 14.6, 4.3 NSA 0.939 NSA 0.996 NSA 0.979 * Migration rates used by Gómez-Alpizar et al. (1)

! 18!

Table S8. Root population inferred using RAS sequences from Mexico and Andean samples from this study. Migration rates were symmetric for all runs. Seed 1 Seed 2 Seed 3 Seed 4 Simulations 1×106 1×106 1×107 1×107 θ M Root Prob. Root Prob. Root Prob. Root Prob. 1.5 0.5 Mexico 0.541 Mexico 0.584 Andes 0.510 Mexico 0.521 5.0 Mexico 0.638 Mexico 0.703 Mexico 0.564 Mexico 0.583 10.0 Andes 0.723 Mexico 0.975 Mexico 0.612 Andes 0.794 2.0 0.5 Andes 0.676 Andes 0.638 Mexico 0.513 Mexico 0.514 5.0 Mexico 0.883 Andes 0.614 Mexico 0.590 Andes 0.794 10.0 Mexico 0.714 Andes 0.688 Mexico 0.663 Mexico 0.939

! 19!

Supporting Figures Figure S1. Structure plots for K from 2 to 10 for the Andes (with and without admixture) and Mexico samples of P. infestans. Figure S2. Delta K values obtained from Structure Harvester for (A) Mexico and (B) Andes (with admixture) samples of P. infestans. Note the highest delta K value corresponds to K=4 for Mexico and K=2 for the Andes. Figure S3. Root state posterior probabilities for each location for (A) Clade 1c and (B) P. infestans. Independent analysis of each locus produced high probabilities for the root of both Clade 1c and P. infestans in Mexico (green bar). Figure S4. Maximum clade credibility genealogy for the β-tubulin locus. Colors of branches indicate the most probable geographic origin of each lineage. Figure S5. Maximum clade credibility genealogy for the PITG_11126 locus. Colors of branches indicate the most probable geographic origin of each lineage. Figure S6. Maximum clade credibility genealogy for the RAS locus. Colors of branches indicate the most probable geographic origin of each lineage. Figure S7. Maximum clade credibility genealogy for the Trp1 locus. Colors of branches indicate the most probable geographic origin of each lineage. Figure S8. Maximum clade credibility phylogeny of Phytophthora Clade 1c from all four loci. Colors of branches indicate the most probable geographic origin of each lineage and taxon colors represent species (purple: P. phaseoli; blue: P. ipomoeae; green: P. mirabilis; red: P. andina; black: P. infestans). Figure S9. Scenarios used to test evolutionary relationships among Andes lineages EC-1, PE-3, and PE-7 using DIYABC. Present day populations are at the base of the tree schematic. Ancestral relationships among these populations are represented by lines intersecting in the past, with the vertex of the schematic representing the most recent common ancestor of all samples. Horizontal lines indicate admixture events between the ancestral populations connected by the horizontal line. Change in line thickness indicates a potential change in population size, such that all branches in scenarios D, E, and F had independently estimated population sizes. The scenarios in which PE-3 and PE-7 diverged more recently from each other than from EC-1 produced high posterior probabilities. Figure S10. Scenarios used to test evolutionary relationships of EC-1, PE-3, and PE-7 lineages to US-1 in the Andes and to the Toluca, Mexico population. The 15 scenarios were compared against each other, using three independent sets of scenarios with EC-1, PE-3, and PE-7, in turn, substituted for population ‘X’. Present day populations are at the base of the tree schematic. Ancestral relationships among these populations are represented by lines intersecting in the past, with the vertex of the schematic representing the most recent common ancestor of all samples.

! 20!

Horizontal lines indicate admixture events between the ancestral populations connected by the horizontal line. Potential changes in population size over time are indicated by changes in line thickness, but line thickness is not proportional to population size. The dashed line represents an unsampled population that has contributed to the genetic variation observed in sampled populations.

A. Andes (No admixture model) C. Mexico (Admixture model)B. Andes (Admixture model)

A. Mexico

B. Andes

RAS/iRASTrp1B-tubulin PITG_11126

0.0

0.2

0.4

0.6

0.8

CO EC ES HU MX NL PE PO SA SW UK USA VT CO EC ES HU MX NL PE PO SA SW UK USA VT CO EC ES HU MX NL PE PO SA SW UK USA VT CO EC ES HU MX NL PE PO SA SW UK USA VT

Pos

terio

r Pro

babi

lity

of R

oot S

tate

0.0

0.2

0.4

0.6

0.8

CO EC ES HU MX NL PE PO SA SW UK USA VT CO EC ES HU MX NL PE PO SA SW UK USA VT CO EC ES HU MX NL PE PO SA SW UK USA VT CO EC ES HU MX NL PE PO SA SW UK USA VT

Location (Country)

A) Clade 1c

B) P. infestans

Pos

terio

r Pro

babi

lity

of R

oot S

tate

RAS/iRASTrp1B-tubulin PITG_11126

Figure S3. Root state posterior probability values for each location after 300 millions MCMC generations by locus for (A) Clade 1c species and (B) P. infestans. The posterior probability for a Mexico root (greenbar) is higher than for any of the other countries for each locus independently and both datasets.

location.states

Colombia

Ecuador

Estonia

Hungary

Mexico

Netherlands

Peru

Poland

South Africa

Sweden

United Kingdom

United States

Vietnam

0.04 substitutions per site

PiVT03a

PiMX4

5a

PmMX09a

PiUK02a

PaEC14a

PaEC

15a

PiES0

2b

PiMX04b

PiES0

4a

PiUK06a

PiCO02a

PaEC09a

PiEC03a

PiPE06a

PiMX2

9a

PiMX07a

PiMX18b

PiUK03b

PiMX65a

PiHU02a

PiMX05b

PoMX03b

PiMX06a

PiUS11a

PiPE08b

PiMX14a

PmMX12b

PpUS01b

PiMX50a

PaEC01a

PiPE02b

PaEC08b

PiEC02a

PiMX72a

PiMX46a

PiUK07a

PiMX17a

PiUK04b

PiEC06a

PaEC12b

PiCO01a

PiPO

01b

PaEC06a

PiES0

4b

PiPE1

1b

PmMX10b

PiUK02b

PmMX03a

PiMX15b

PiMX41a

PiNL02b

b12

EPi

P

PiMX18a

PiES0

3a

PaEC04b

PiES0

5b

PaEC08a

PiPE1

2b

PiCO01c

PiMX30b

PaEC07a

PiPE22b

PiMX0

5a

PiMX45b

PiPE0

6b

PiCO05c

PiSW

01b

PiPE1

2a

PiPE2

5a

PiMX66a

PiSA01a

PiPE2

9a

PaEC17b

PiUK04a

PiMX16a

PaEC02aPiPO02a

PaEC10b

PaEC10a

PoMX04a

PiEC10b

b90K

UiP

PiMX23b

PiMX67a

PiMX42b

PmMX01b

PiVT04bPiVT03b

PiUK08a

PiEC1

1a

PiSW

02a

PaEC03b

PiUK03a

PpUS02b

PoMX06b

PiPE1

4a

PiPO02b

PiMX14b

PiMX41b

PiMX21b

PiMX28a

PiPE2

8a

PiMX49a

PiCO03a

PiPE05b

PaEC07b

PaEC17a

PiPE2

4b

PiMX4

3a

PaEC09b

PmMX05b

PiUK10b

PiEC07b

PiUK09a

PiEC07a

PiMX1

2a

PmMX03b

PiMX11a

PiEC08b

PiMX29b

PiMX07b

PiES0

5a

PiMX20a

PiPE1

4b

PaEC16b

PiMX13a

PiMX71b

PiVT02a

PiMX12b

PiUK05b

PiMX4

3b

PiEC11b

PiUK01a

PiMX63a

PaEC05a

PaEC14b

PiEC1

2a

PpUS02a

PiMX26b

PiUK10a

PpUS01a

PoMX01a

b80K

UiP

PiMX26a

PaEC12a

PiMX06b

PmMX04b

PaEC11b

PiEC10a

PiMX13b

PmMX11a

PoMX08bPiM

X17b

PiMX1

5a

PiMX27b

PaEC13b

PoMX05a

PiPE07b

PiPE2

4a

PiEC03b

PmMX11b

PiMX6

4a

PiMX10b

PmMX05a

PiVT01a

PiMX4

7b

PiEC13a

PiUK05a

PiUS17a

PmMX04a

PiPO01a

PiMX64b

PiPE0

2a

PaEC

03a

PiMX16bPiMX50b

PaEC01b

PiPE13a

PiNL01b

PiPE1

0b

PiMX25b

PiPE

20b

PiMX22b

PaEC11a

PiMX40b

PiMX01a

PiEC14a

PmMX12a

PiES0

1a

PiMX4

4a PiUS11b

PiMX11b

PaEC02b

PiMX1

0a

PiMX22aPiMX20b

PiUS08b

PiMX27a

PiMX24a

PiMX72b

PaPE02b

PiPE08a

PiPE0

4a

PiMX46b

PiSA01b

PiPE2

7a

PiSW

01a

PoMX05b

PiMX30a

PiM

X61a

PmMX10a

PiEC14b

PiPE0

1a

PoMX02b

PiMX62b

PiMX4

4b

PoMX03a

PiMX02b

PiPE2

3a

PoMX01b

PiNL01a

PiPE0

5a

PiMX71a

PiCO05b

PiMX0

4a

PiMX42a

PiPE04b

PiPE2

6b

PiVT02b

PiPE2

6aPiP

E27b

PaEC15b

PiUK06b

PiES0

2a

a20

XMi

P

PaEC05b

PiMX62a

PiUK

01b

PiCO04a

PaEC13a

PiMX4

0a

PiMX48b

PiMX28b

PaEC16a

PmMX09b

PoMX06a

a01E

PiP

PiUS08a

PiPE1

3b

PiMX65b

PiUS17b

PmMX02a

PiMX63b

PiPE0

9a

PiPE0

1bPiEC12b

PiMX49b

PiMX2

3a

PiCO05a

PaEC06bPiMX03b

PiMX48a

PiPE

29b

PiMX2

5a

PiES01b

PiPE

28b

PiMX01b

PiPE2

0a

PiCO03b

PiMX47a

PiPE2

5b

PmMX02b

PoMX08a

PiUK07b

PiEC0

6b

PiM

X67b

PiSW

02b

PiPE0

7a

a12

EPi

P

PiUS12a

PmMX06b

PiEC13b

PiMX21a

PiCO02b

PiEC02b

PaPE01b

PoMX02a

PiMX66b

PiES0

3b

PiVT01b

PiPE2

2a

PmMX06a

PiPE09b

PiMX19a

PoMX04b

PiMX61b

PiCO01b

PaEC04a

PiVT04a

PaPE

01a

PiPE2

3b

PiEC08a

PmMX01a

PiMX24b

PaPE

02a

PiMX19b

PiNL02a

a11E

PiP

PiMX0

3a

PiHU02b

PiUS12b

PiCO04b

MX

EC

MX

MX

MX

MX

US

Figure S4. Maximum clade credibility genealogy for the β-tubulin locus. Colors of branches indicate the most probable geographic origin of each lineage.

PiUK05a

PiEC

06b

PmMX07a

PaPE02a

PmMX04b

PiMX17b

PaEC11a

PiPE08a

PiEC

14b

PaEC06b

PiMX49b

PaEC11b

PiMX17a

PiUK08c

PiMX12a

PiUK08a

PmMX01b

PiMX24b

PaEC10b

PiMX21

a

PiMX64b

PiUK05b

PiPE14aPi

EC13

b

PiPE09b

PiUK10a

PiUK02b

PiEC

10a

PiUK07a

PiMX43b

PoMX05

a

PiHU02a

PiPE03b

PiPE14b

PiEC08b

PpUS01a

PiUK10b

PiES04b

PiPE

25b

PoMX04a

PiPE13b

PiMX23a

PaEC10a

PaPE01b

PiMX25a

PiM

X46b

PaEC12a

PiM

X22a

PiMX01

a

PiPE02a

PiVT

02b

PaEC13b

PiVT

03b

PmMX06b

PiM

X30b

PiMX25b

PiPE03a

PiMX42b

PiMX07b

PiPE06a

PiVT04a

PmMX04a

a60

CEaP

PiPE

01a

PmMX05a

PmMX05b

PmMX02b

PiEC02a

PmMX09a

PiMX05b

a11EPiP

PiPE10a

PiMX72

a

PoMX 08a

PiMX06a

PiM

X18a

PiCO02b

PiUK08b

PiSA

01a

PiPE07a PaEC07b

PpUS01b

PiMX26b

PiPO

01a

PiM

X01b

PiES01b

PmMX10b

PiPE09a

PiEC

13a

PiUS17a

PaEC09aPiCO04a

PiM

X48a

PiES

04a

PoMX01b

PiMX41

a

PiPE

13a

PiMX41b

PiPE

12a

PaEC17a

b81X

MiP

PiPE26b

PiPE

27b

PiUK06c

PiUS17b

PiSW02a

PiES

05a

PiMX62

a

PoMX02b

PiCO04b

PiPE11b

PaEC13a

PmMX10a

PiMX16b

PiPE05aPaEC12b

PiVT

03a

PiM

X29a

PiES02a

PiMX26a

b41X

MiP

PiM

X50b

PiEC07b

PiCO02aPiEC12a

PoMX03a

PiEC01b

PiM

X28b

PiPE04b

PiPE12b

PiMX40b

PiMX27a

PiMX49a

PaEC01a

PiM

X44b

PoMX04

b

PiPE

23b

PoMX06a

PaEC15a

PmMX11a

PiUK01bPiSW01b

PiMX13a

PiEC01a

PaEC14b

PiVT01a

PaEC04a

PiUK03b

PiMX46a

PiPE28a

PiES01a

PmMX01aPmMX03a

PiUK04b

PiMX10b

PpUS02b

PiPE29a

PiMX24a

PoMX02a

PoMX03b

PiPE23a

PiM

X15b

PiC

O05

a

PaEC04b

PiM

X40a

PiSW02b

PiMX07a

PiMX42

a

PaEC01b

PmMX02a

PiPE

22a

PiMX47b

PiEC03b

PiPE06b

PiM

X66a

PiSW01a

PiUS08a

PiMX71b

PaEC16b

PaEC

07a

PmMX12b

PiMX62b

PiPE

25a

PiUS12a

PiM

X02a

PiEC

11b

PiM

X64a

PiMX66

b

PiMX14a

PiUK02a

PiPE02b

PaEC02a

PiEC14a

PiMX21b

PiMX30a

PiCO03a

PiCO03b

PaEC16a

PaEC08b

PiMX61b

a61XMiP

PiMX0

5a

PiPE04a

PiPE14c

PiEC

10b

PaEC15b

PiUS12b

PiUS11a

PiEC

12b

PiUS08b

PmMX12a

PaEC03b

PiPE24a

PiM

X44a

PiPE

20b

PaEC09b

PiMX50

a

PiMX45a

PiUK09a

PiPE27a

PiUK01a

PiPE

21b

PiMX19a

PaEC02b

PmMX09b

PiMX65b

PaEC05b

PiUK03a

PiPE20a

PiPE05b

PiM

X03a

PiUK06a

PiPE07bPaEC17b

PiVT02a

PaEC05a

PaPE

01a

PiMX04

a

PiMX61a

PaEC08a

PiMX06b

PoMX05b

PiPO02b

PiMX20b

b91XMiP

PiMX29b

PiNL01a

PiEC02b

PiES02b

PiVT

01b PiM

X11a

PiMX22b

PiMX03b

PiPE29b

PiUK06b PoMX06

b

PiHU02b

PiES

03a

PiMX67b

PmMX07b

PiVT

04b

PiCO01a

PiMX11b

PiNL01b

PiEC02c

PoMX01a

PoMX08 b

a30CEaP

PiM

X04b

PiM

X65a

PiEC11a

PiM

X12b

PiPO

02a

PiMX47

a

PiCO01b

PiMX43a

PmMX11b

PiCO05b

PiPE

22b

PiSA

01b

PiEC07a

PiPE01b

PiEC03a

PiPE21a

PpUS02a

PiPE08b

PiUK09b

PiUK07b

PiMX48b

PiMX71a

PiM

X63b

PiM

X63a

PiMX02b

PiMX28aPi

PE24

b

PiUS11b

PiES03b

PiM

X67a

PiMX72b

PmMX06a

PiMX15a

PiMX23b

PiMX13b

PiPE26a

PiNL02b

PiPE28b

PaEC14a

PiEC

06a

PiMX27b

PiNL02a

PiEC08a

PaPE02b

PiM

X45b

PiPE10b

PiM

X10a

PiES05b

PiUK04a

PiPO01b

PmMX03b

PiM

X20a

location.states

Colombia

Ecuador

Estonia

Hungary

Mexico

Netherlands

Peru

Poland

South Africa

Sweden

United Kingdom

United States

Vietnam

MX

MX

US

MX

MX

2.0E-4 substitutions per site

EC

Figure S5. Maximum clade credibility genealogy for the PITG_11126 locus. Colors of branches indicate the most probable geographic origin of each lineage.

2.0E-4 substitutions per site

PiMX6

6a

PiMX4

2b

PiMX2

4b

PiMX4

4a

PiSA0

1b

PmMX01b

PaEC08a

PiMX72b

PiEC02a

PiMX2

5a

PiES0

5b

PmMX06b

PiMX18b

PaEC10a

PaEC07b

PiVT03b

PaEC04a

PiCO0

3a

PiMX07b

PiMX49b

PoMX05b

P10KUib

PiMX6

1b

a10E

PiP

PiMX03b

PiPE04a

PaEC05a

PiMX6

3a

PiPE21a

PiMX3

0b

PiCO0

2b

PiPE21b

PiNL02b

PmMX05a

PiPE0

9a

PiMX27b

PiMX40a

PiMX21a

PiES0

5a

PiEC10b

PaEC08b

PmMX03b

b20S

EiP

PiMX4

2a

PoMX02a

PmMX09a

PaPE02b

PaEC06b

PiES0

2a

PiMX67a

PiCO04a

PaEC15a

PiM

X24a

PiPE26b

PiMX48b

PiPE26a

PiVT04a

PiEC03a

PiPE03b

PaEC01b

PiMX4

4b

PaEC13a

PiPO01a

PiUS08b

PiMX06b

PiEC01b

PiPO02a

PiES04b

PiMX7

2a

PiPE2

5a

PiCO03b

PaEC09b

PoMX04bPiPO01b

PiUK09a

PoMX01b

PiMX29a

PiMX1

2b

PiEC03b

PiSW01a

PiPE04b

PiMX16ab60KUiP

PoMX03a

PiMX17a

PiEC14b

PiES03b

PiMX03a

PiUK

01a

PiPE01b

PiPE09b

PmMX04a

PiMX0

4b

PiPE0

5a

PaEC10b

PiMX46b

PiMX19b

PaEC13b

PaEC06a

PiES0

1b

PiPE20b

PiMX19a

PiMX25b

PiM

X02a

PoMX05a

PiMX2

6a

PiPE07b

PiMX61a

PaEC15b

PiMX18a

PiEC06b

PiMX27a

PiCO01b

PiPE1

4a

PiPE22b

a11EPiP

PiMX11b

PiVT01b

PoMX06a

PiPE12b

PiMX14a

PiVT03a

PiEC13b

PiEC12a

PiUS08a

PiSA0

1a

PaEC04b

PiMX1

7b

PiM

X15a

PoMX07b

PiMX5

0a

PpUS01b

PmMX11a

PiUK1

0a

PiMX4

3a

PiVT01a PiPE1

2a

PiPE2

4a

PiES0

4a PiSW01b

PiMX20b

PiMX05a

PiUK06a

PiPE14b

PiMX64aPiUK07a

PaEC01a

PmMX09b

PiMX6

6b

PiVT04b

PiMX62b

PaEC02b

PiM

X41a

PaEC14a

PiVT02b

PaEC09a

PiPE02b

PiEC08a

PiEC11a

PiUK05a

PmMX08aPiPO02b

PiMX65b

PoMX04a

PaEC11b

PiMX49a

PiUK

02b

PiMX48a

PiMX4

5a

PiPE0

3a

PmMX11b

PiEC13a

PmMX10a

PiM

X28a

PmMX03a

PmMX07a

PiUK

04b

PaEC02a

PiEC02b

PiCO

05b

PiPE08a

PiPE2

2a

PiMX62aPoMX07a

PiEC06a

PiPE05b

PiMX67b

PiMX0

6a

PiEC08b

PmMX05b

PaEC11a

PiPE08b

PiMX4

1b

PiEC11b

PiMX16b

PiPE25b

PiMX71a

PmMX07b

PaEC07a

PaEC05b

PiMX1

4b

PiMX6

4b

PaEC12b

PmMX04b

PmMX01a

PiEC10a

PiVT02a

PiMX13a

PiMX46a

PiMX28b

PiPE2

7a

PaPE02a

PiMX47b

PiCO0

5a

PiMX11a

PiNL02a

PiPE0

2a

PiUS12a

PiMX4

5b

PiUK02a

PiPE1

3a

PiMX15b

PiMX21b

PiPE29b

PiNL01a

PmMX08b

PaEC03a

PiPE06b

PiPE11b

PiES0

3a

PiCO0

1a

PiMX47a

PiMX26b

PaPE01b

PiUK10b

PiMX2

2b

PiMX02b

PiMX22a

PiUK0

4a

PiMX01a

PiPE28b

PoMX02b

PiPE23b

PiEC14a

PiES01a

PiMX40b

PiPE10b

PiCO0

4b

PiPE28a

PiPE06a

PiPE27bPiPE24b

PiMX13b

PmMX10b

b90KUiP

PiMX4

3bPpUS01a

PiPE13b

PiMX01bPiMX63b

PiUK

05b

PaEC12a

PaEC14b

PiMX50b

PaEC03b

PiPE29a

PoMX03b

PiEC01a

PiUK07b

PiMX30a

PiCO02a

PiPE10a

PiMX29b

PiMX12a

PoMX06b

PaPE01a

PiNL01b

PiMX65a

PoMX01a

PiMX7

1b

PiMX20a

PiMX0

4a

PiPE07a

PmMX06a

PiSW02a

PiMX07a

PiPE2

0a

PiPE2

3a

PiSW02b

PiEC12b

PiUS12b

PiMX0

5b

MX US

EC

MX

MX

location.states

Colombia

Ecuador

Estonia

Hungary

Mexico

Netherlands

Peru

Poland

South Africa

Sweden

United Kingdom

United States

Vietnam

MX

Figure S6. Maximum clade credibility genealogy for the RAS locus. Colors of branches indicate the most probable geographic origin of each lineage.

PiCO

03a

PiEC10b

PiPE13a

PiEC07b

PiMX15a

PiMX41b

PiMX07b

PiPE

06a

PiMX48a

PiM

X62b

PiMX42b

PiPE02a

PoMX05b

PiUK04a

PiM

X18a

PiMX19b

PmMX12b

PiSA01b

PiMX03b

PiEC11b

PiPE24b

PiPE

21aPa

PE02

a

PiEC14a

PoMX04a

PiPE12b

PmMX03a

PiPE22b

PaEC17b

PiCO05b

PoMX06b

PaEC13a

PiPE12aPiPE24a

PpUS01b

PiMX05b

PiMX25b

PiEC07a

PiCO02b

PiPE

10a

PiMX44b

PmMX02b

PiSW02b

PaEC16

a

PiMX12a

PiMX50b

PiMX61b

PmMX06a

PiMX62a

PiM

X67b

PiMX02a

PiMX14a

PmMX02a

PmMX01a

PiM

X43a

PiM

X26b

PmMX07a

PiVT0

3a

PiPE09a

PpUS02a

PiMX11a

PiES

05b

PiU

K06a

PiPO02b

PmMX08b

PaEC04a

PiVT04b

PiPE29a

PiMX64b

PoMX06a

PpUS01a

PiVT03b

PaEC15b

PiPE

26a

PiPE25b

PiMX63a

PaEC08a

PiEC02b

PiMX10a

PaEC13b

PiCO04b

PaEC10b

PiMX45b

PmMX10a

PoMX03b

PiPE

22a

PiU

K03b

PiES04a

PiMX01b

PiU

S11a

PaEC05b

PiU

K01a

PiMX15b

PaEC01b

PiMX30a

PaEC

14a

PiMX25a

PiES

03a

PiVT04a

PiMX46a

PoMX05a

PaEC03b

PaEC

07a

PmMX04a

PiUK08a

PiMX06a

PiMX17b

PiMX46b

PiPE05a

PiEC01b

PiMX29b

a71SUiP

PiCO

05a

PiH

U02

a

PiMX03a

PiCO

03b

PiMX14b Pi

UK0

9a

PaEC08b

PiEC06b

PiEC08b

PiPE29b

PiNL02b

PiMX06b

PiEC

01a

PiMX29a

PaEC02b

a32X

MiP

PiM

X22b

PiPE03b

PiPE

01a

PmMX05a

PaEC11b

PoMX02b

PiUK10b

PiPE

05b

PiHU02

b

PiEC13a

PiMX20a

PiEC03b

PiPE13b

PiMX49a

a30KUiP

PoMX01a

PiPE26

b

PiPO01b

PoMX08a

PiMX44a

PiES

03b

PiU

K10a

PaEC04b

PiMX47b

PmMX09b

PiMX48b

PiM

X63b

PmMX11a

PiPE10b

PiMX50a

PaEC17a

PiMX64a

PiES02a

PiEC13b

PmMX12a

PiES

01b

PiMX45a

PaEC05a

PiM

X30b

PmMX03b

PiNL01b

PiMX23b

PiPE21b PiM

X49b

PpUS02b

PiPE

04a

PiMX28b

PiEC12a

PiCO

04a

PiPO

01a

PiUS12b

PaEC01

a

PiPE

23a

PiUS08b

PiUK08b

PiVT01b

PiMX17a

PiMX21b

PiMX07a

PaEC

02a

PiCO

02a

PaEC09b

PiEC

06a

PiPE27a

PiUK05b

PiMX02b

PiEC

03a

PiM

X72b

PiCO

01a

PaEC03

a

PoMX03a

PiEC

02a

PiMX61a

PiMX40a

PiMX65b

PiMX43b

PmMX01b

PaEC

10a

PiPE07a

PiPE20b

PiPO02a

PmMX04b

PiPE28aPiPE23b

PiPE09

b

PmMX11b

PaEC06

a

PiMX27b

PiCO

01b

PaEC14b

PaPE01b

PiMX22a

PiMX20b

PiMX16b

PiMX28a

a50X

MiP

PiMX66a

PaEC06b

PiMX71a

PiMX21a

PiMX04b

PiM

X13b

PiVT02b

PiMX47a

PiMX66b

PiVT0

2a

PmMX07b

PiPE

11a

PiMX24a

PiES05a

PiUK04b

PiPE07b

PiMX16a

PmMX06b

PiU

S12a

PiPE04b

PiSW01a

PiMX27a

PiUK09b

PiUS17b

PiEC11a

PiVT0

1a

PiM

X13a

PiMX26a

PiUK07b

PiPE25a

PiMX42a

PiPE20a

PaEC07b

PiMX19a

PiPE

01b

PiUK01b

PiEC14b

PaEC11a

PiEC08a

PaEC

12a

PiES

02b

PaEC15a

PiU

K07a

PaPE02b

PiMX04a

PiMX41a

PiUK05a

PiPE11b

PiM

X18b

PoMX04b

PiUK06b

PiSW02a

PiPE02b

PiEC10a

PaEC

09a

PiM

X10b

PaEC16b

PoMX01b

PiMX40b

PiU

K02a

PiMX24b

PiNL02a

PiM

X12b

PaEC12b

PiMX65a

PiMX67a

PmMX10b

a80S

UiP

PiNL01a

PmMX05b

PiEC12b

PiMX72a

PiPE06b

PiES

04b

PiUS11b

PiPE27b

PiPE28b

PiES

01a

PiMX01a

PiMX11b

PaPE

01a

PiSA01a

PiUK

02b

PmMX09a

PiM

X71b

PoMX02a

PiPE

03a

PiSW01b

MX

US

MX

MX

MX

MX

location.states

Colombia

Ecuador

Estonia

Hungary

Mexico

Netherlands

Peru

Poland

South Africa

Sweden

United Kingdom

United States

Vietnam

4.0E-5 substitutions per site

Figure S7. Maximum clade credibility genealogy for the Trp1 locus. Colors of branches indicate the most probable geographic origin of each lineage.

3.0E-4

PiUK08a

PiMX49b

PpUS02a

PiMX28

a

PiPE14c

PmMX07b

PiES0

2bPiM

X63b

PiM

X40a

PiUS08b

PiPE22a

PiVT02a

PaPE01b

PaEC16b

PaEC

16a

PiMX2

5a

PiSW

01b

PiMX45b

PiCO0

5c

PiPE13a

PiPE26a

PaPE01a

PiM

X49a

PiCO0

1c PiES04b

PiEC03b

PpUS01a

PiPE06a

PoMX01b

PiUS1

1b

PaEC

13a

PiMX12a

PiVT03b

PiEC02a

PmMX05b

PmMX01a

PiMX72

b

PiES03b

PiMX29b

PaEC

15a

PmMX02b

PiUK04a

PiMX27a

PiMX24a

PiSA0

1a

PaEC03b

PiUK06b

PiPE24a

PiPE25b

PaEC17b

PoMX02a

PiPE29a

PaEC04b

PiPO02a

PiCO03a

PmMX06a

PiMX0

6a

PiMX28b

PiMX7

1a

PaEC02b

PiEC11b

PmMX12b

PiEC10b

PiMX13b

PiMX27b

PiEC0

3a

PiMX07b

PiPE1

2b

PiMX11a

PoMX05a

PiMX6

2a

PiPE27b

PiMX42b

PiMX26b

PiUS1

1a

PiM

X62b

PiCO0

5a

PiUS1

7a

PiUK10a

PaPE02a

PoMX07a

PiMX61a

PmMX04a

PiES04a

PpUS01b

PiMX01

b

PiMX22b

40OCiP

a

PiMX50a

PiPE11a

PaEC

12a

PmMX02a

PmMX09a

PiUS12b

PaEC08b

PmMX01b

PiMX15a

PiUK01a

PmMX04b

PaEC04a

PmMX03aPmMX11a

PiMX0

2a

PoMX04b

PiMX42a

PiUS17b

PiPO

01b

PiMX2

4b

PaEC11a

PiNL0

2a

PiCO01a

PoMX06a

PiPE24b

PiUK03b

PiCO

02a

PiPE0

3a

PiEC13aPoMX04a

PiMX19b

PpUS02b

PaEC

05a

PiMX06bPiMX02b

PiEC02b

PiPE28a

PiSW

02b

PiPE0

2a

PiMX20a

PiUK

01b

PiMX04b

PiMX2

5b

PiPE22b

PiMX48b

PiPE21a

PiCO02b

PiMX48a

PiSA01b

PiMX46a

PaEC11b

PiMX47b

PmMX03b

PiCO04b

PiPE26b

PiMX41a

PiMX23a

PiMX67a

PiPE23a

b30X

MiP

PiUK07a

PiCO03b

PiMX40b

PiM

X61bPiM

X72a

PiMX15b

PiVT02b

PiPE20b

PiES0

1b

PiVT03a

PiEC14a

PiMX21a

PiPE28b

PaEC13b

PiMX05b

PiMX41b

PiUK09a

PaEC17a

PiMX17b

PiMX6

6b

PiM

X65a

PiMX07a

PiPE20a

PiMX04a

PiPE1

4b

PiMX1

4b

PiMX26

a

PiMX45a

PiPE1

1b

PoMX03bPiVT04a

PiPE0

8a

PiUK08b

PaEC

14a

PiEC13b

PiMX05a

PiMX18b

PiPE04b

PiPE05b

PiMX44a

PaEC15b

PiUS12a

PiMX1

8aPiPE14a

PaEC

08a

PmMX09b

b20OPiP

PiMX3

0a

PiMX44b

PiMX46b

PiNL01a

PiNL02b

PaEC10b

PaEC01b

PiES05b

PiMX13aPiVT01b

PiEC07a

PiES05a

PiMX10a

PiPO01a

PiES01a

PiMX16

a

PiUK

08c

PiUK03a

PiMX1

7a

PiVT01a

PaEC07b

PiPE10b

PiEC14b

PiEC0

8a

PiMX6

6a

b34

XMi

P

PiPE0

5a

PmMX07a

PiM

X65b

PaEC03a

PmMX10b

PiMX64a

PiES03a

PiMX03aPiUK06c

PiPE29b

PiPE08b

PiEC06b

PiPE0

1b

PiCO01b

PoMX02bPiEC06a

PoMX03a

PiPE21b

PiEC08b

PiMX19a

PiVT04b

PiMX10

b

PiPE1

3b

PiMX3

0b

PiEC07b

PaEC01a

PiMX29a

PiUK06a

PiEC01aPiMX22a

PiMX16b

PiHU02b

PaPE02b

PiUS0

8a

PmMX10a

PiES02a

PiPE27a

PiUK0

2a

b60E

PiP

PaEC

09a

PiPE01a

PiMX6

3a

PiPE10a

b90KUiP

PoMX07b

b40K

UiP

PiEC01

b

PaEC

06a

PaEC10a

PiMX23

b

PiUK02b

PiPE0

9a

PiMX20

b

PiPE09b

PaEC02a

PiPE07b

PoMX01a

PmMX05a

PiMX71

b

PiMX67b

PaEC14b

PiPE02b

PiUK10b

PoMX06bPoMX08b

PiPE12a

PiNL01b

PiMX64b

PiMX01a

a74X

MiP

PiSW02a

PiUK05a

PiPE04a

PiMX11b

PiMX12b

PiMX50b

PiPE07a

PaEC09b

PiPE03b

PiEC02c

PoMX05b

PiUK07b

PiMX14a

PiEC1

2a

PmMX12a

PiEC10a

PmMX06b

PiPE23b

PiM

X43a

PoMX08a

PiUK

05b

PaEC05b

PaEC

07a

PiMX2

1b

PiPE25a

PiCO05b

PaEC12b

PiEC11a

PiHU02a

PaEC06b

PmMX11b

PiEC12b

PiSW01a

MX

EC

MX MX

MX

MX

MX

MX

location.states

Colombia

Ecuador

Estonia

Hungary

Mexico

Netherlands

Peru

Poland

South Africa

Sweden

United Kingdom

United States

Vietnam

EC1 PE3PE7 PE3 EC1PE7 EC1 PE7PE3

EC1 PE3PE7 PE3 EC1PE7 EC1 PE7PE3

0.9 [0.7, 1.1] 21.9 [19.7, 24.0] 0.9 [0.7, 1.2]

1.4 [1.1, 1.7] 73.5 [71.1, 75.8] 1.4 [1.1, 1.7]

A. B. C.

D. E. F.

Toluca US1X X US1Toluca Toluca XUS1

Toluca X US1US1 Toluca XToluca US1 X

X TolucaUS1Toluca US1 XToluca US1X

X US1 TolucaToluca X US1Toluca US1 X

X US1 TolucaUS1 X TolucaX Toluca US1

1. 2. 3.

4. 5. 6.

7. 8. 9.

10. 11. 12.

13. 14. 15.