taxonomy of mexican diploid wild potatoes (solanum … cabrera and spooner... · 2005. 6. 13. ·...

TAXONOMY OF MEXICAN DIPLOID WILD POTATOES (SOLANUM

SECT. PETOTA): MORPHOLOGICAL AND MICROSATELLITE DATA

Sabina I. Lara-Cabrera and David M. Spooner

ABSTRACT. Solanum sect. Petota, the potato and its wild relatives, contains about 200 wild species distributed from the southwestern United States to central Argentinaand adjacent Chile. Most species grow in the Andes, but the United States, Mexico, andCentral America contain about 30 diploid (2n = 24) taxa, as well as tetraploid (2n = 48) and hexaploid (2n = 72) ones. Chloroplast DNA restriction site data showed13 of these 30 taxa to form a clade containing only diploid species, but there was low resolution within the clade. Some of these 13 taxa are similar morphologically, and we questioned whether they were valid species. We analyzedthese and related taxa with morphological and microsatellite data. Morphological datashowed extensive overlap of putative species-specific characters, but most species couldbe supported by multivariate techniques, except S. brachistotrichium and perhaps S. stenophyllidium. Previously mapped nuclear microsatellite markers, developed in S. tuberosum, also were explored for help to define species but performed poorly, showing that these microsatellites have reduced phylogeneticutility to analyze the United States, Mexican, and Central American diploid wild potato species.

Key words: microsatellite, potato, simple sequence repeats, Solanum sect. Petota, SSR.

9

200

A FESTSCHRIFT FOR WILLIAM G. D’ARCY

Solanum sect. Petota Dumort., the potatoand its wild relatives, contains about 200wild species distributed from the south-western United States to central

Argentina and adjacent Chile (Hawkes, 1990;Spooner & Hijmans, 2001). Most species grow inthe Andes, but the United States, Mexico, andCentral America contain about 30 diploid taxa(2n = 24), as well as tetraploid (2n = 48) and hexa-ploid (2n = 72) ones. The United States, Mexico,and Central American diploids are classified inSolanum series Bulbocastana, Morelliformia,Pinnatisecta, Polyadenia, and Tuberosa (Hawkes,1990); see Table 1 for authorities of series, species,and subspecies. For simplicity, we refer to speciesfrom this region as Mexican or primarily Mexicanspecies. Some Mexican diploids appear very simi-lar morphologically and led us to questionwhether they were valid species. In this paper weexplore the use of morphology and microsatel-lites (or simple sequence repeats, SSRs) to test thevalidity of these primarily Mexican diploidspecies.

Hawkes (1989, 1990) divided Solanum sect.Petota into 21 series. Two of these series are non-tuber-bearing, and 19 are tuber-bearing. Hawkes(1989, 1990) placed the two non-tuber-bearingseries in subsection Estolonifera Hawkes, and theremaining 19 tuber-bearing series in subsectionPotatoe Don of Solanum. He divided subsectionPotatoe into two superseries, Stellata Hawkesand Rotata Hawkes, based on the corolla shape.Hawkes (1990) and Hawkes and Jackson (1992)proposed a phylogeny of subsection Potatoe ofSolanum based on data from morphology, bio-geography, ploidy level, and Endosperm BalanceNumber. The Endosperm Balance Number (EBN)relates to a strong isolating mechanism inSolanum sect. Petota and explains success or fail-ure of intra- and interspecific crosses, due to thefunctioning or breakdown of the endospermafter fertilization. The EBNs are hypotheticalgenetic factors independent of ploidy and empir-ically determined, relative to crossing with stan-dard test EBN crossers of known EBN. In potato,these are 2x(1EBN), 2x(2EBN), 4x(2EBN), 4x(4EBN),

and 6x(4EBN). Species can cross within EBN, irre-spective of ploidy (Hanneman, 1994).

Hawkes (1990) and Hawkes and Jackson (1992)informally designated subsets of Solanum specieswithin each superseries with more “primitive” or“advanced” morphological characters, providingfour groups, primitive Stellata, advanced Stellata,primitive Rotata, and advanced Rotata. In thistheory the Mexican diploids are primitive Stellata(in ser. Bulbocastana, Morelliformia, Pinnatisecta,and Polyadenia). All Solanum species in theseseries are 2x(1EBN). Representatives of theseprimitive Stellata in Solanum migrated to SouthAmerica and remained primitive Stellata (ser.Circaeifolia, Commersoniana, Lignicaulia,Olmosiana). These then evolved to advancedStellata, then to primitive Rotata, and finally toadvanced Rotata. The species in Mexico that areadvanced Rotata (S. verrucosum of ser. Tuberosa,and members of ser. Conicibaccata, Demissa, andLongipedicellata) either remigrated back fromSouth America, or are products of hybridizationbetween these remigrants and primitive indige-nous Mexican diploids. For the sake of simplicity,all Solanum species of primitive Stellata in theUnited States, Mexico, and Guatemala will herebe referred to as “primitive Mexican diploids.”

Spooner et al. (1991) investigated the Hawkes(1990) Stellata–Rotata hypothesis using chloro-plast DNA (cpDNA) restriction site analysis inSolanum series Bulbocastana, Conicibaccata,Demissa, Longipedicellata, Morelliformia,Pinnatisecta, Polyadenia, and Tuberosa. Theirresults gave partial support to the Hawkes (1989,1990) and Hawkes and Jackson (1992) hypothe-ses, by supporting most primitive Mexicandiploids as basal, and other Mexican species asadvanced. Some putative primitive Stellata fromSouth America, however, such as Solanum cir-caeifolium (ser. Circaeifolia), were later supportedas derived by further cpDNA restriction site stud-ies (Spooner & Sytsma, 1992; Spooner & Castillo,1997).

Two other cpDNA restriction site studies investi-gated relationships of Mexican potatoes.Spooner and Sytsma (1992) produced a clado-

TAXONOMY OF MEXICAN DIPLOID WILD POTATOES (SOLANUM SECT. PETOTA)

201

1 21

Bu

lboc

asta

naSo

lanu

m b

ulbo

cast

anum

Dun

al

2751

98

Haw

kes

1595

MEX

ICO

. Méx

ico:

Cer

ro d

e Co

coti

tlán

(Ryd

b.) H

awke

ssu

bsp.

bul

boca

stan

um(IB

UG

, MEX

U, P

TIS)

2 21

S.

bul

boca

stan

um27

5184

H

awke

s 15

81

MEX

ICO

. Fed

eral

Dis

tric

t: P

edre

gal d

e Sa

nsu

bsp.

bul

boca

stan

umÁ

ngel

, Ciu

dad

Uni

vers

itar

ia (I

BUG

, IEB

, M

EXU

, PTI

S, W

IS)

3 23

S.

bul

boca

stan

um27

5194

H

awke

s 15

91

MEX

ICO

. Oax

aca:

rui

ns o

f M

onte

albá

n su

bsp.

bul

boca

stan

um(IB

UG

, PTI

S)

4 14

S.

bul

boca

stan

um49

8224

O

choa

141

62

MEX

ICO

. Mic

hoac

án: L

as P

alm

itas

su

bsp.

dol

icho

phyl

lum

(Bit

ter)

Haw

kes

(CIP

, IBU

G, K

, PTI

S, W

IS)

5 21

S.

bul

boca

stan

um54

5752

Ta

rn e

t al

. 244

M

EXIC

O. M

éxic

o: E

of

Coat

epéc

Har

inas

, su

bsp.

dol

icho

phyl

lum

betw

een

San

Fran

cisc

o an

d Po

rfir

io D

íaz

(B, K

, IBU

G, P

TIS)

6 17

S.

bul

boca

stan

um25

5516

G

raha

m 3

00B

MEX

ICO

. Jal

isco

: Cui

dad

Guz

mán

(PTI

S)

subs

p. d

olic

hoph

yllu

m

7 28

S.

bul

boca

stan

um55

8379

Sp

oone

r et

al.

MEX

ICO

. Chi

apas

: 9.4

km

S o

f Rt

. 190

su

bsp.

par

titu

m(C

orre

ll) H

awke

s42

24(f

rom

nea

r Te

opis

ca),

to L

as R

osas

(IB

UG

, IN

IFA

P, P

TIS,

WIS

)

8 29

S.

bul

boca

stan

um27

5200

H

awke

s 17

96

GU

ATE

MA

LA. H

uehu

eten

ango

: rd.

fro

m

subs

p. p

arti

tum

Hue

huet

enan

go t

o Q

ueza

lten

ango

, 5

km S

of

Man

laca

tanc

ito

(C, K

, PTI

S)

Acc

essi

on

Map

Seri

es p

lace

men

t

Tax

on

PI3

Col

lect

or

Loc

alit

yN

umbe

r1Lo

calit

y2by

Haw

kes

(199

0)

TA

BL

E1.

Loca

tion

, ser

ies

affi

liati

on w

ithi

n So

lanu

m, P

I num

ber,

and

colle

ctor

, of

the

acce

ssio

ns e

xam

ined

in t

his

stud

y. H

erba

rium

vou

cher

s ar

e de

posi

ted

at P

TIS,

alon

g w

ith

othe

r kn

own

herb

aria

of

depo

siti

on in

the

loca

lity

colu

mn

(CIP

= In

tern

atio

nal P

otat

o Ce

nter

in L

ima,

Per

u, IN

IFA

P =

the

Mex

ican

Nat

iona

lPo

tato

Pro

gram

in T

oluc

a, M

exic

o).1

Acc

essi

on n

umbe

r co

rres

pond

s to

Fig

s. 2

, 4. 2

The

map

loca

lity

corr

espo

nds

to F

ig. 1

. SA

ref

ers

to S

outh

Am

eric

a an

dU

ref

ers

to lo

calit

ies

wit

h un

know

n lo

cati

on b

eyon

d th

e co

untr

y of

ori

gin.

3U

nite

d St

ates

Dep

artm

ent

of A

gric

ultu

re P

lant

Intr

oduc

tion

Num

bers

.


Acc

essi

on

Map

Seri

es p

lace

men

t

Tax

on

PI3

Col

lect

or

Loc

alit

yN

umbe

r1Lo

calit

y2by

Haw

kes

(199

0)

9 27

S.

cla

rum

Corr

ell

5583

82Sp

oone

r et

al.

MEX

ICO

. Chi

apas

: 1.1

km

N o

f to

wn

4216

squa

re o

f El

Por

veni

r on

rd.

to

Silt

epec

guez

(IBU

G, I

NIF

AP,

PTI

S)

10

23

Mor

ellif

orm

iaH

awke

s S.

mor

ellif

orm

eBi

tter

& M

ünch

27

5220

H

awke

s 16

48

MEX

ICO

. Ver

acru

z: K

m 3

05.5

fro

m M

éxic

o on

the

rd.

fro

m L

as V

igas

to

Jala

pa,

Mal

país

de

La J

oya

(C, K

, P, P

TIS)

11

16

S. m

orel

lifor

me

27

5218

H

awke

s 16

13

MEX

ICO

. Méx

ico:

44.

5 km

fro

m t

he m

ain

Tolu

ca–M

orel

ia R

d., t

owar

ds V

alle

de

Brav

o fr

om T

oluc

a (C

, PTI

S, U

S)

12

16

S. m

orel

lifor

me

49

8003

Ta

rn e

t al

. 48

MEX

ICO

. Méx

ico:

abo

ut 1

8.5

km t

owar

ds

Valle

de

Brav

o fr

om T

oluc

a–Te

mas

calt

epec

, Hw

y. 1

34 v

ia R

anch

o Sa

n Ra

món

(PTI

S)

13

5 Pi

nnat

isec

ta(R

ydb.

) S.

bra

chis

totr

ichi

um(B

itte

r) R

ydb.

32

0265

H

awke

s 12

34

MEX

ICO

. Chi

huah

ua: M

ajal

ca, 4

0 m

i. H

awke

sN

W o

f Ch

ihua

hua,

nea

r th

e vi

llage

(C

, IBU

G, P

TIS,

US)

145

S. b

rach

isto

tric

hium

49

8217

O

choa

142

06

MEX

ICO

. Chi

huah

ua: R

t. C

hihu

ahua

–Ci

udad

Juá

rez

(CIP

, IBU

G, P

TIS,

US,

WIS

)

15

6 S.

bra

chis

totr

ichi

um

4979

93

Tarn

et

al. 1

3 M

EXIC

O. C

hihu

ahua

: La

Aur

ora

at 4

1 km

, al

ong

rd. t

o Sa

n Ju

anit

o fr

om t

he L

a Ju

nta

to Y

epac

hic

Hw

y. (P

TIS)

16

8 S.

bra

chis

totr

ichi

um

5458

32

Tarn

et

al. 2

07

MEX

ICO

. Agu

asca

lient

es: f

rom

Rin

cón

de

Ram

os o

n H

wy.

45,

30

km a

long

the

tra

ck

tow

ards

El C

hiqu

ihui

tillo

, Ran

cho

Tier

ra

Colo

rada

(BR,

C, K

, MEX

U, P

TIS)

TA

BL

E1

CO

NT

INU

ED.

202

203


17

8 S.

bra

chis

totr

ichi

um

5458

15

Tarn

et

al. 2

15

MEX

ICO

. Agu

asca

lient

es: H

wy.

70,

abo

ut

33 k

m f

rom

Agu

asca

lient

es, 8

.2 k

m a

long

th

e tr

ack

past

Milp

illas

de

Arr

iba

tow

ards

Po

trer

o (P

TIS)

18

8 S.

bra

chis

totr

ichi

um

2555

27

Gra

ham

346

X

MEX

ICO

. Agu

asca

lient

es: b

etw

een

348

Agu

asca

lient

es a

nd C

alvi

llo (P

TIS)

19

22

S. c

ardi

ophy

llum

Lind

l. 34

7759

Ta

rn 2

41D

M

EXIC

O. P

uebl

a: le

ft s

ide

of T

ehua

cán–

subs

p. c

ardi

ophy

llum

Hua

juap

an d

e Le

ón r

d. o

n th

e Pu

ebla

sid

e of

bor

der,

Hw

y. 1

25, a

lmos

t at

bor

der

of

Pueb

la–O

axac

a st

ates

(PTI

S)

20

U

S. c

ardi

ophy

llum

2830

62

Gra

ham

s.n

. M

EXIC

O (P

TIS)

subs

p. c

ardi

ophy

llum

21

U

S. c

ardi

ophy

llum

2830

63

Gra

ham

s.n

. M

EXIC

O (P

TIS)

su

bsp.

car

diop

hyllu

m

22

11

S. c

ardi

ophy

llum

2792

72

Haw

kes

1458

M

EXIC

O. A

guas

calie

ntes

: 7 m

i. fr

om

subs

p. e

hren

berg

iiBi

tter

Agu

asca

lient

es o

n th

e rd

. to

Lore

to,

1.25

mi.

from

the

mai

n A

guas

calie

ntes

to

Zaca

teca

s H

wy.

(C, K

, PTI

S, U

S)

23

12

S. c

ardi

ophy

llum

1847

62

Haw

kes

1086

M

EXIC

O. Q

ueré

taro

: San

Jua

n de

l Río

, su

bsp.

ehr

enbe

rgii

SW o

f th

e to

wn

(ECO

N, P

TIS)

24

8 S.

car

diop

hyllu

msu

bsp.

ehr

enbe

rgii

3412

31

Com

mon

wea

lthM

EXIC

O. J

alis

co: K

m 2

01 f

rom

Po

tato

G

uada

laja

ra o

n th

e rd

. to

San

Luis

Pot

osí

Colle

ctio

n 23

02(P

TIS)

25

12

S. c

ardi

ophy

llum

subs

p. e

hren

berg

ii18

6548

H

awke

s 11

00

MEX

ICO

. Zac

atec

as: 7

26.5

km

fro

m M

éxic

o to

war

ds Z

acat

ecas

(LL,

PTI

S)

26

9 S.

car

diop

hyllu

msu

bsp.

ehr

enbe

rgii

2555

20

Gra

ham

371

M

EXIC

O. S

an L

uis

Poto

sí(P

TIS)

27

12

S. c

ardi

ophy

llum

subs

p. e

hren

berg

ii27

5213

H

awke

s 14

28

MEX

ICO

. Que

réta

ro: 3

km

SSE

of

tow

n (C

, K, P

TIS,

US)

Acc

essi

on

Map

Seri

es p

lace

men

t

Tax

on

PI3

Col

lect

or

Loc

alit

yN

umbe

r1Lo

calit

y2by

Haw

kes

(199

0)


Acc

essi

on

Map

Seri

es p

lace

men

t

Tax

on

PI3

Col

lect

or

Loc

alit

yN

umbe

r1Lo

calit

y2by

Haw

kes

(199

0)

28

22

S. h

into

niiC

orre

ll

Spoo

ner

et a

l. M

EXIC

O. M

éxic

o: 6

.5 k

m S

W o

f 40

33Te

mas

calt

epec

–Val

le d

e Br

avo

rd.,

on r

d. t

o Sa

n Pe

dro

Tena

yac,

on

S si

de

of r

d., c

a. 5

0 m

dow

nstr

eam

of

brid

ge

over

rd.

(IN

IFA

P, P

TIS)

29

2 S.

jam

esii

Torr

. 45

8425

U

gent

&

U.S

.A. A

rizo

na: A

pach

e Co

., H

wy.

180

, Ru

hde

16-7

83.

2 km

S o

f Ea

ger

at t

he ju

ncti

on w

ith

Rt. 1

30 (W

IS, P

TIS)

30

2 S.

jam

esii

56

4051

Sa

las

et a

l. 24

U

.S.A

. Ari

zona

: Apa

che

Co.,

near

Nel

son

rese

rvoi

r, tw

o m

i. S

on R

t. 6

66 t

hen

E 1

mi.

on r

d. 2

75 (P

TIS)

31

3 S.

jam

esii

5640

49

Sala

s et

al.

21

U.S

.A. N

ew M

exic

o: C

atro

n Co

., Re

serv

e vi

cini

ty, a

t ab

out

6 m

i. N

E of

Ara

gon

on

Hw

y. 1

2, a

t 33

mi.

mar

ker

(fro

m H

wy.

180

) (P

TIS)

32

1 S.

jam

esii

2751

69

Haw

kes

1176

U

.S.A

. New

Mex

ico:

Gra

nt C

o., S

end

of

the

Big

Burr

o M

ount

ains

, Silv

er C

ity,

17

mi.

from

Lor

dsbu

rg (R

t. 1

80) (

K, P

TIS)

33

1 S.

jam

esii

45

8423

U

gent

&

U.S

.A. N

ew M

exic

o: G

rant

Co.

, Hw

y. 9

0,

Ruhd

e 7-

7843

.5 k

m S

W o

f Si

lver

Cit

y, G

ila N

atio

nal

Fore

st (P

TIS,

WIS

)

34

2 S.

jam

esii

27

5172

H

awke

s 12

07

U.S

.A. A

rizo

na: C

ochi

se C

o., C

hiri

cahu

a M

ount

ains

, Mon

ita

Cano

n, ju

st b

elow

Fa

raw

ay R

anch

at

the

edge

of

Nat

iona

l M

onum

ent

(K, P

TIS)

TA

BL

E1

CO

NT

INU

ED.

204

205


35

20

S. ×

mic

hoac

anum

(Bit

ter)

55

8497

Sp

oone

r et

al.

MEX

ICO

. Mic

hoac

án: a

t km

21

mar

ker

Rydb

. (pu

tati

ve o

rigi

n is

40

77al

ong

Rt. 1

20 S

of

Mor

elia

, on

W s

ide

of

S. b

ulbo

cast

anum

×S.

pin

nati

sect

um)

Rt. 1

20 (I

NIF

AP,

PTI

S)

36

7 S.

nay

arit

ense

(Bit

ter)

Ryd

b.

5458

27

Tarn

et

al.

MEX

ICO

. Nay

arit

: nea

r St

a. A

nita

, abo

ut

270A

1.5

km N

E of

Sta

. Ter

esa

(PTI

S)

37

8 S.

nay

arit

ense

54

5820

Ta

rn e

t al

. M

EXIC

O. Z

acat

ecas

: Hw

y. 5

4, 1

0 km

SW

23

4Aof

Jal

pa, 2

6.3

km a

long

the

tra

ck t

owar

d Tl

alte

nang

o (K

, PTI

S)

38

8 S.

nay

arit

ense

54

5825

Ta

rn e

t al

. 225

M

EXIC

O. Z

acat

ecas

: Hw

y. 7

0, 1

2 km

NW

of

Jal

pa, 1

8 km

alo

ng t

he t

rack

SE

tow

ard

Tlac

hich

ila (K

, PTI

S)

39

13

S. p

inna

tise

ctum

Dun

al

2752

34

Haw

kes

1456

M

EXIC

O. J

alis

co: 1

5 m

i. fr

om L

eón

on t

he

rd. t

o A

guas

calie

ntes

(C, I

BUG

, K, M

EXU

, PT

IS, U

S)

40

11

S. p

inna

tise

ctum

18

4764

H

awke

s 10

93

MEX

ICO

. Gua

naju

ato:

Leó

n, 2

km

SE

of t

he

cem

ent

fact

ory

(BR,

IBU

G, K

, LL,

MEX

U,

MPU

, NY,

PTI

S, W

AG

)

41

14

S. p

inna

tise

ctum

27

5233

H

awke

s 14

55

MEX

ICO

. Gua

naju

ato:

2 m

i. fr

om S

iláo

on

the

rd. t

o G

uana

juat

o, E

l Cas

tillo

(C, I

BUG

, K

, MEX

U, P

TIS)

42

13

S. p

inna

tise

ctum

27

5236

H

awke

s 15

05

MEX

ICO

. Jal

isco

: 29

mi.

from

Gua

dala

jara

on

the

rd.

to

Mex

ico

City

, hac

iend

a de

H

uejo

tita

n (C

, IBU

G, K

, MEX

U, P

TIS,

US)

43

12

S. p

inna

tise

ctum

27

5230

H

awke

s 14

24

MEX

ICO

. Que

réta

ro: 3

km

SSE

of

tow

n, b

y th

e sm

all r

eser

voir

(IBU

G, K

, MEX

U, P

TIS)

44

20

S. p

inna

tise

ctum

19

0115

Ba

ldw

in14

374

MEX

ICO

. Mic

hoac

án: M

orel

ia (L

L, N

A, P

TIS)

45

13

S. b

ulbo

cast

anum

Rodr

ígue

z M

EXIC

O. J

alis

co: S

ierr

a de

La

Prim

aver

a ×

S.ca

rdio

phyl

lum

puta

tive

hyb

rid

2592

(PTI

S)

Acc

essi

on

Map

Seri

es p

lace

men

t

Tax

on

PI3

Col

lect

or

Loc

alit

yN

umbe

r1Lo

calit

y2by

Haw

kes

(199

0)


Acc

essi

on

Map

Seri

es p

lace

men

t

Tax

on

PI3

Col

lect

or

Loc

alit

yN

umbe

r1Lo

calit

y2by

Haw

kes

(199

0)

46

11

S. ×

sam

buci

num

Rydb

. 59

5478

Ro

dríg

uez

MEX

ICO

. Que

réta

ro: f

oot-

hills

of

“El B

uey”

(p

utat

ive

orig

in is

S. e

hren

berg

ii25

63m

ount

ain,

on

junc

tion

of

road

s of

×

S. p

inna

tise

ctum

)Q

ueré

taro

–San

Lui

s Po

tosí

and

Mex

ico

City

–San

Lui

s Po

tosí

(F, I

BUG

, MIC

H, M

O,

NY,

PTI

S)

47

18

S.×s

ambu

cinu

m60

4209

Ro

dríg

uez

MEX

ICO

. Gua

naju

ato:

La

Purí

sim

a,

2565

mun

icip

alit

y of

San

Die

go d

e la

Uni

ón,

rd. f

rom

Que

réta

ro c

ity

to S

an L

uis

Poto

sí

(MIC

H, P

TIS)

48

10

S. s

teno

phyl

lidiu

mBi

tter

55

8460

Sp

oone

r et

al.

MEX

ICO

. Jal

isco

: rd.

fro

m N

sid

e of

41

04G

uada

laja

ra, p

ast

side

rd.

to

San

Fran

cisc

o Te

sist

án t

o Sa

n Cr

istó

bal d

e la

Bar

ranc

a,

on W

sid

e of

rd.

, 20.

5 km

(by

post

ed k

m

sign

s), N

of

Gua

dala

jara

, abo

ut 0

.5 k

m S

of

Mir

ador

(IN

IFA

P, P

TIS)

49

16

S. t

arni

iHaw

kes

& H

jert

. 49

8048

Ta

rn e

t al

. 79

MEX

ICO

. Hid

algo

: fro

m L

as T

ranc

as,

Hw

y. 8

5, Z

imap

an t

o Ja

cala

(K, P

TIS)

50

16

S. t

arni

i 54

5808

Ta

rn e

t al

. 257

M

EXIC

O. H

idal

go: t

urni

ng o

ff H

wy.

85

(Zim

apan

/Jac

ala)

at

Las

Tran

cas,

go

6.4

km

alon

g rd

. tow

ards

Nic

olás

Flo

res

(C, F

, K,

MEX

U, P

TIS,

WA

G)

51

15

S. t

arni

i 54

5742

Ta

rn e

t al

. 56

MEX

ICO

. Ver

acru

z: a

bout

14

km S

of

Hua

yaco

cotl

a, n

ear

Vib

orill

os, o

n rd

. to

Tula

ncin

go (K

, PTI

S)

52

16

S. t

arni

i 49

8043

Ta

rn e

t al

. 57

MEX

ICO

. Hid

algo

: 28

km S

of

Hua

yaco

cotl

a, 8

km

bef

ore

Agu

a Bl

anca

(C

, K, M

EXU

, PTI

S)

TA

BL

E1

CO

NT

INU

ED.

206

207


53

16

S. t

arni

i 57

0641

H

jert

ing

7361

MEX

ICO

. Hid

algo

: 20

km f

rom

Agu

a Bl

anca

tow

ards

Hua

yaco

cotl

a, n

ear

Palo

Be

ndit

o (P

TIS)

54

17

S. t

rifi

dum

Corr

ell

2555

42

Cana

da

MEX

ICO

. Mic

hoac

án: K

m 4

8 on

the

O

ttaw

a 46

0Ca

rapa

n–U

ruap

an H

wy.

(PTI

S)

55

18

S. t

rifi

dum

55

8478

Sp

oone

r et

al.

MEX

ICO

. Mic

hoac

án: o

n rd

. fro

m R

t. 1

20

4283

NW

to

Cher

an, 2

.5 k

m S

E of

Ser

ina

(INIF

AP,

PTI

S, W

IS)

56

17

S. t

rifi

dum

55

8480

Sp

oone

r et

al.

MEX

ICO

. Mic

hoac

án: a

long

mic

row

ave

4089

Bto

wer

rd.

off

of

dirt

rd.

beg

inni

ng a

bout

5

km N

E of

Cap

acua

ro, 5

.9 k

m f

rom

be

ginn

ing

of t

his

mic

row

ave

tow

er r

d.

(INIF

AP,

PTI

S)

57

17

S. t

rifi

dum

28

3104

H

awke

s 15

41

MEX

ICO

. Jal

isco

: NE

slop

e of

Nev

ado

de

Colim

a, 1

3 km

fro

m t

urn

to T

olim

án o

n th

e rd

. fro

m C

iuda

d G

uzm

án, L

os A

lpes

(IB

UG

, K, M

EXU

, PTI

S)

58

U

S. t

rifi

dum

29

2213

U

gent

574

5 M

EXIC

O (P

TIS)

59

26

Poly

aden

iaBu

kaso

v S.

lest

eriH

awke

s &

Hje

rt.

5584

34

Spoo

ner

et a

l.

MEX

ICO

. Oax

aca:

on

W s

ide

of R

t. 1

25,

4155

3.4

km S

of

Rt. 1

90, j

ust

S of

San

Jua

n Te

posc

olul

a (IN

IFA

P, P

TIS,

WIS

)

60

26

S. le

ster

i 44

2694

H

awke

s et

al.

M

EXIC

O. O

axac

a: 2

5.8

km S

of

Mia

huat

lán

1714

on t

he O

axac

a–Pu

erto

Áng

el r

d. (K

, PTI

S)

61

26

S. le

ster

i 55

8435

Sp

oone

r et

al.

M

EXIC

O. O

axac

a: a

long

E s

ide

of R

t. 1

75,

4177

at k

m 1

34.1

, abo

ut 5

0 m

off

rd.

; up

a sl

ope,

S o

f M

iahu

atlá

n de

Por

firi

o D

íaz

(INIF

AP,

PTI

S, W

IS)

Acc

essi

on

Map

Seri

es p

lace

men

t

Tax

on

PI3

Col

lect

or

Loc

alit

yN

umbe

r1Lo

calit

y2by

Haw

kes

(199

0)


Acc

essi

on

Map

Seri

es p

lace

men

t

Tax

on

PI3

Col

lect

or

Loc

alit

yN

umbe

r1Lo

calit

y2by

Haw

kes

(199

0)

62

22

S. p

olya

deni

umG

reen

m.

4980

36

Tarn

et

al. 6

5 M

EXIC

O. H

idal

go: H

wy.

85,

Zim

apan

to

Tam

zunc

hale

, at

Las

Tran

cas,

abo

ut 8

km

E

at J

ague

y on

the

tra

ck t

o N

icol

ás F

lore

s (P

TIS)

63

13

S. p

olya

deni

um

5584

50

Spoo

ner

et a

l.

MEX

ICO

. Jal

isco

: on

mic

row

ave

tow

er r

d.

4137

to C

erro

Gra

nde,

SE

of S

ante

Fé,

1.1

km

do

wnh

ill o

f to

p of

tow

n (IN

IFA

P, P

TIS,

WIS

)

64

24

S. p

olya

deni

um

1617

28

Corr

ell 1

4374

M

EXIC

O. M

icho

acán

: nea

r M

atug

eo

(ECO

N, G

AT,

MEX

U, P

TIS)

65

25

S. p

olya

deni

um

3477

69

Tarn

& G

ómez

M

EXIC

O. P

uebl

a: N

ew M

éxic

o–V

erac

ruz,

22

1to

ll rd

., at

km

208

.5 (K

, PTI

S)

66

19

S. p

olya

deni

um

3477

70

Tarn

& G

ómez

MEX

ICO

. Ver

acru

z: n

ear

Puer

to d

el A

ire,

23

4on

old

Ori

zaba

–Méx

ico

rd.,

Rt. 1

50,

betw

een

Puer

to a

nd V

erac

ruz–

Pueb

la

stat

e bo

unda

ry (K

, PTI

S)

67

SA

Circ

aeif

olia

Haw

kes

S. c

irca

eifo

lium

Bitt

er

4981

88

Haw

kes

et a

l.BO

LIV

IA. C

ocha

bam

ba: A

yopa

ya, n

ear

6581

Inde

pend

enci

a, Ic

hupa

mpa

(PTI

S)

68

SA

Com

mer

soni

aBu

kaso

vS.

com

mer

soni

Dun

al

5667

51

Hof

fman

s.n

. A

RGEN

TIN

A (P

TIS)

69

SA

Coni

ciba

ccat

aBi

tter

S.

san

tolla

lae

Varg

as

4733

72

Haw

kes

et a

l.

PERU

. Cuz

co: P

auca

rtam

bo, P

illah

uata

51

03(P

TIS)

70

SA

Lign

icau

liaH

awke

s S.

lign

icau

leVa

rgas

27

5273

E.

Bau

r

PERU

. Cuz

co: C

alca

, Pis

ac (P

TIS)

So

rtim

ent

1884

71

2 Lo

ngip

edic

ella

taS.

fen

dler

isub

sp. f

endl

eriH

awke

s 27

5163

H

awke

s 11

80

U.S

.A. A

rizo

na: C

ochi

se C

o., C

hiri

cahu

a Bu

kaso

vM

ount

ains

, Rus

tler

Par

k (K

, PTI

S)

72

16

Tube

rosa

(Ryd

b.)

S. v

erru

cosu

mSc

hltd

l.27

5260

H

awke

s 16

58

MEX

ICO

. Hid

algo

(AA

, BM

, C, K

, PTI

S)H

awke

s

TA

BL

E1

CO

NT

INU

ED.

208


209

gram that divided the primitive Mexican diploidspecies into these clades in Solanum: (1) seriesPinnatisecta, with internested series Polyadenia,Bulbocastana (in part), and series Morelliformia;(2) S. bulbocastanum and S. cardiophyllum; and(3) S. verrucosum and all Mexican polyploidspecies. The second clade was unexpectedbecause all prior hypotheses suggested that S.bulbocastanum and S. cardiophyllum would asso-ciate with the first clade. Solanum verrucosum,on the other hand, was long recognized by mor-phological and crossing data to truly be distinctand to belong in the third clade. Rodríguez andSpooner (1997) reanalyzed the Solanum bulbo-castanum and S. cardiophyllum clade includingmore accessions and all the subspecies of eachspecies. The resulting tree maintained all the sub-species in the same clade except for all the acces-sions of S. cardiophyllum subsp. ehrenbergii. Thissubspecies was placed in the basal Mexicandiploid clade, most closely related to S. brachis-totrichium and S. stenophyllidium.

The basal relationship of the primitive Stellatawithin Solanum sect. Petota, including S. cir-caeifolium of series Circaeifolia, was furtherinvestigated by Kardolus (1998) using amplifiedfragment length polymorphisms (AFLP). Solanumcircaeifolium was shown to be primitive by AFLPsand nested with the Mexican diploids, unlikecpDNA results (Spooner & Castillo, 1997), sug-gesting a “chloroplast capture” event, known tobe common throughout angiosperms (Rieseberg& Brunsfeld, 1992). Its primitive position, howev-er, matched the Stellata/Rotata hypothesis ofHawkes (1989, 1990) and Hawkes and Jackson(1992).

Microsatellites are hypervariable molecular mark-ers (Tautz, 1989; Powell et al., 1996), consisting oftandem repeats of 1 to 6 bp (Provan et al., 1996b;Rubinsztein et al., 1999; Bichara et al., 2000). Theyhave been used in cultivar identification(Smulders et al., 1997; Lin & Walker, 1998),

germplasm analysis (Powell et al., 1996), thedevelopment of quantitative trait loci (QTL)(Arens et al., 1995; Areshchenkova & Ganal, 1999;Winter et al., 1999; Lorieux et al., 2000), to detectgenetic polymorphisms (Russell et al., 1997; Naderet al., 1999; Teulat et al., 2000), and in phyloge-netic studies (Kostia et al., 2000; Petren et al.,1999; Clisson et al., 2000; Raker & Spooner, 2002).In potato they have been used mainly for cultivaridentification (Milbourne et al., 1997; Schneider &Douches, 1997), to develop linkage groups(Milbourne et al., 1998), in germplasm identifica-tion (Provan et al., 1996b), in analysis of intra-spe-cific somatic hybrids (Provan et al., 1996a), and ingenetic analysis of anther-derived haploid pota-toes (Veilleux et al., 1995).

Milbourne et al. (1998) developed microsatelliteprimers and mapped them to all 12 linkagegroups of cultivated potato, S. tuberosum. Wewished to test their use to investigate the validityand interspecific relationships of the primitiveMexican diploids and other primitive andadvanced potatoes, including members ofHawkes’s (1990) primitive Stellata. Raker andSpooner (2002) showed these primers to be of useto distinguish the subspecies of S. tuberosum, butthis was the very germplasm base from whichthese primers were developed. We needed thesedata for our ongoing floristic studies of Solanumsect. Petota in North America, Mexico, andCentral America. This included investigations ofhybrid origins for some of these species. Thesehybrids and putative origins by Hawkes (1990) areS. ×michoacanum (S. bulbocastanum × S. pinnati-sectum), and S. ×sambucinum (S. cardiophyllumsubsp. ehrenbergii × S. pinnatisectum). In addi-tion, Rodríguez and Vargas-P. (1994) postulated anew unnamed hybrid of S. bulbocastanum × S.cardiophyllum (no subspecies designated foreither species). We include South American prim-itive Stellata species to test the Hawkes (1989,1990) and Jackson and Hawkes (1992) hypothesesof relationships in Solanum.

210


MATERIALS AND METHODSPLANT MATERIALMore than one accession per taxon was analyzed(when available) from as many geographicallydispersed sites as possible (Table 1, Fig. 1). Theplants were grown in the greenhouse, becausemany of the primitive Mexican diploids do notthrive outdoors in Wisconsin. An average of threeplants per accession of the Mexican diploids weregrown, and one accession per series from theSouth American diploids. Solanum stenophyllidi-um had only one accession available as seeds, andwe analyzed two separate sets of three plants perthe same accession for morphological analysis.

MORPHOLOGY AND DATA ANALYSISMorphological characters were measured whenthe plants were in full flower (Table 2). Significantdifferences in character states for quantitativecharacters were studied by the Tukey-Kramer

test, and qualitative characters by the LikelihoodRatio Chi square test and Pearson Chi square testin JMP statistical software v. 3.1.5 (SAS InstituteInc., 1998). Phenetic analyses were performedwith NTSYS-pc (Rohlf, 1992). An operational tax-onomic unit (OTU) was the average of three indi-viduals per accession. Data were standardized(STAND), and similarity matrices (in SIMINT), aver-age taxonomic distance (DIST), Euclidean distance(EUCLID), Manhattan distance (MANHAT), andproduct-moment correlation (CORR) were gener-ated. Clustering was performed with theunweighted pair-group method, UPGMA (Sokal& Sneath, 1963), and Neighbor Joining, NJ (Saitou& Nei, 1987). Cophenetic correlation coefficients(COPH, in MXCOMP) were used to measure thedistortion between similarity matrices and theresultant four phenograms. Stepwise discriminateanalyses (SDA) were performed by SAS v.7 (SASInstitute Inc., 1998) using Stepwise discriminateanalysis (STEPDISC).

Figure 1. Distribution of wild potato accessions used in this study from the United States, Mexico, and Guatemala. Accessionsfrom South America are not mapped. Numbers correspond to generalized map localities in Table 1.


211

TABLE 2.Morphological characters measured. All values are in cm, unless otherwise indicated. The 35 of 56statistically significant characters (P = 0.05, SAS Institute Inc., 1998) distinguishing all taxa (see text) aremarked with an asterisk (*).

STEM CHARACTERS1. Stem pubescence length (mm)*. 2. Stem pubescence posture: pointing up (1), pointing down (2),

pointing out (3)*. 3. Stem diameter (mm)*. 4. Stem wings: present (1), absent (2). 5. Stem shape: angular(1), rounded (2)*. 6. Pubescence type: glabrescent (1), pubescent (2)*.

LEAF CHARACTERS7. Leaf width*. 8. Ratio: leaf length/width*. 9. Number of leaflet pairs. 10. Leaflet position: alternate

(1), opposite (2). 11. Widest leaflet pair: terminal (1), most distal lateral leaflet pair (2), second most distal lateral leaflet pair (3)*. 12. Interstitial leaflets: present (1), absent (2)*. 13. Terminal leaflet width*. 14. Ratio:length/width of terminal leaflet blade*. 15. Length from widest point of terminal leaflet blade to base ofblade*. 16. Ratio: length from widest point of terminal leaflet blade to base of blade/length of terminalleaflet blade. 17. Petiolule length of terminal leaflet*. 18. Most distal lateral leaflet width*. 19. Ratio: lengthof most distal lateral leaflet blade/width of most distal lateral leaflet*. 20. Length from widest point of mostdistal lateral leaflet blade to base of blade*. 21. Ratio: length from widest point of most distal lateral leafletblade to base of blade/length of most distal lateral leaflet blade*. 22. Petiolule length of most distal lateralleaflet*. 23. Second most distal lateral leaflet width*. 24. Ratio: length of second most distal lateral leafletblade/width of second most distal lateral leaflet*. 25. Length from widest point of second most distal later-al leaflet blade to base of blade*. 26. Ratio: length from widest point of second most distal lateral leafletblade to base of blade/length of second most distal lateral leaflet blade. 27. Petiolule length of second mostdistal lateral leaflet*. 28. Abaxial leaf pubescence length (mm)*. 29. Abaxial leaf pubescence posture: point-ing up (1), pointing down (2), pointing out (3). 30. Lateral leaflet decurrency: decurrent (1), not decurrent(2)*. 31. Petiole wings: present (1), absent (2)*. 32. Petiole length*. 33. Shape of pseudostipular leaflet: pinnulate (1), lunate (2)*. 34. Spicy odor: present (1), absent (2)*.

FLORAL CHARACTERS (see Spooner & van den Berg, 1992, for illustrations of characters 49–52).35. Peduncle length. 36. Length from the base to the articulation of the pedicel*. 37. Ratio: pedicel

length/length from base to the articulation of the pedicel. 38. Length of calyx (mm). 39. Ratio: length ofcalyx lobe/length of calyx*. 40. Posture of calyx lobe: straight (1), bent over (2), curled around (3). 41.Symmetry of calyx: symmetrical (1), asymmetrical (2). 42. Length of calyx acumen (mm)*. 43. Ratio: lengthof calyx lobe/length of calyx acumen (mm). 44. Calyx color: green (1), purple (2), green and purple (3). 45.Position of calyx and corolla: alternate (1), opposite (2). 46. Style exsertion (mm). 47. Length of anther(mm). 48. Length of anther filament (mm). 49. Radius of the corolla measured from the center of the corolla to the tip of the corolla lobes (mm). 50. Ratio: Radius of the corolla measured from the center ofthe corolla to the tip of the corolla lobes/distance between the center of the flower and the junction ofthe corolla lobes (mm)*. 51. Width of the corolla lobe measured at the base of the corolla junction (mm)*.52. Ratio: width of the corolla lobe measured at the base of the corolla junction/length from a line drawnacross widest point of corolla lobes (mm). 53. Primary color of the abaxial surface of the corolla: white (1),creamy white (2), purple (3)*. 54. Secondary color of the abaxial surface of the corolla: white (1), creamywhite (2), purple (3). 55. Posture of style: erect (1), curved (2).

FRUIT CHARACTERS56. Fruit shape: globose (1), conical (2)*.

212


We used stepwise discriminant analysis to discov-er characters that best distinguished the ninemost phenetically similar primitive Mexicandiploid species (Solanum brachistotrichum, S. car-diophyllum subsp. ehrenbergii, S. hintonii, S. jamesii, S. ×michoacanum, S. nayaritense, S. stenophyllidium, S. trifidum, S. tarnii), with allprocedures as above.

DNA EXTRACTION, AMPLIFICATION, AND DNAFRAGMENT ANALYSISDNA was extracted using the protocol in Rakerand Spooner (2002). Forty-five mapped primerswere selected from Milbourne et al. (1998) thatwere designed for S. tuberosum and screened forability to amplify a fragment in these species:stm0001, stm0003, stm0007, stm0010, stm0013,stm0014, stm0017, stm0019, stm0024, stm0025,stm0028, stm0032, stm0037, stm0038, stm0051,stm0052, stm1003, stm1005, stm1008, stm1017,stm1020, stm1021, stm1024, stm1025, stm1029,stm1031, stm1041, stm1049, stm1055, stm1058,stm1064, stm1069, stm1100, stm1102, stm1104,stm1106, stm2005, stm2012, stm2013, stm2020,stm2022, stm2028, stm3009, stm3010, stm3016.Forward primers were labeled with fluorescentdyes FAM (blue), HEX (yellow), and TETRA(green), from PE Biosystems, to amplify individualaccessions of all of the species. Reaction condi-tions were optimized, modifying conditions listedin Provan et al. (1996b). Conditions for a 25 µlreaction included 1X PCR Buffer II (Perkin Elmer),1.5 mM MgCl2, 0.2 mM dNTPs, 0.4 µM of eachprimer pair, 1 unit of AmpliTaq Gold® (PerkinElmer), and 10–20 ng of DNA. The annealingtemperature followed Milbourne et al. (1998) or1ºC to 2ºC below (Table 3). All reactions wereamplified in a 9600 PE thermocycler with the fol-lowing times and temperatures: 1 cycle for 10min. at 94°C, 2 min. to anneal temperature, 5min. at 72°C, 29 cycles of 1 min. at 94°C, 45 sec. toanneal, 5 min. at 72°C, ending with a 72°C holdfor 45 min.

The microsatellite products for each individualwere pooled in dye combinations of 3 µl FAM: 6µl TETRA: 9 µl HEX, or in some cases each product

was analyzed separately. PCR products were sentto the Biotechnology Center at the University ofWisconsin-Madison for fragment separation.Products were processed as follows: 1.5 µlaliquots of PCR product were mixed with 1.8 µlsample loading buffer (80% formamide, 10mg/ml dextran blue, 5 mM EDTA at pH 8.0) and0.45 µl of TAMRA 500 molecular weight standard(PE Biosystems). The mixture was heated for 3min. at 95°C and then placed on ice.Approximately 1 µl of the cold samples wasloaded on a 5% LongRanger™ (FMC Bioproducts)polyacrylamide/6M urea gel in a PE Biosystems377XL DNA sequencing machine. Samples wererun at 3000 V with 2400 scans/hour in 36 cm well-to-read plates. Data were collected using a PEBiosystems DNA Sequencer data collection v. 2.0and analyzed with GeneScan v. 2.1 (PEBiosystems). Data were later imported in toGenotyper v. 2.1 PE Applied Biosystems for allelesizing, both automatically and manually. Peakswere scored as allele sizes in bp and manuallytranslated into binary presence/absence.

MICROSATELLITE DATA ANALYSISThe multistate matrix was imported to Microsat v.1.5d (Minch et al., 1997) to produce distancematrices with two models, absolute differenceand delta mu squared (∂µ)2, both of whichassume a step-wise mutation model (SMM)(Goldstein et al., 1995). New mutations followingthis model are obtained by adding or subtractingrepeats one by one (Goldstein et al., 1995). Thesedistance matrices were later imported into PAUP4.0b3a (Swofford, 1998) to build trees withUPGMA (Sokal & Sneath, 1963) and NJ (Saitou & Nei, 1987).

The binary matrix was imported into PAUP 4.0b3a(Swofford, 1998) to produce the distance matrixof Nei and Li (1979) that applies the infinite allelemodel (IAM), where every mutation produces anew allele of potentially infinite size (Kimura &Crow, 1964). The Nei and Li matrix was used tobuild UPGMA (Sokal & Sneath, 1963) and NJ(Saitou & Nei, 1987) trees.

213


Prim

er

Tag

Pr

imer

seq

uenc

eT

arge

ted

repe

atSi

ze r

ange

A

nnea

ling

C

hrom

osom

e Pe

rcen

t (b

p)Te

mp.

(°C

)m

issi

ng d

ata

stm

0003

-f

B

gg

a g

aa t

ca t

aa c

aa c

ca g

(a

c)9

(at)

9 10

1–17

0 48

X

II 4%

stm

0003

-r

aat

tgt

aac

tct

gtg

tg

t g

tg

stm

0007

-f

B

gg

a ca

a g

ct g

tg a

ag t

tt a

t (a

c)9

185–

234

52

XII

58%

stm

0007

-r

aat

tga

gaa

ag

a g

tg t

gt

gtg

stm

0014

-f

G

cag

tct

tca

gcc

cat

ag

g

(gt)

5 (a

t)7

(gt)

10

77–2

01

53

I 18

%

stm

0014

-r

taa

aca

atg

gta

gac

aag

aca

aa

stm

0051

-f

Y

tac

ata

cat

aca

cac

acg

cg

(a

c)7…

(ac)

7 (a

t)4

84–1

16

53

X

23%

stm

0051

-r

ctg

caa

ctt

ata

gcc

tcc

a

stm

1005

-f

B

atg

cct

ctt

acg

aat

aac

tcg

g

(gta

)6

160–

187

59

VIII

27

%

stm

1005

-r

cag

cta

acg

tg

g t

tg g

gg

stm

1017

-f

Gg

ac a

cg t

tc a

cc a

ta a

a (a

tt)5

11

5–13

7 48

IX

21

%

stm

1017

-r

aga

aga

ata

gca

aag

caa

stm

1024

-fB

at

a ca

g g

ac c

tt a

at t

tc c

cc a

a (t

tg)6

12

5–15

2 59

V

III

13%

stm

1024

-r

tca

aaa

ccc

aat

tca

atc

aaa

tc

stm

1029

-f

Y

agg

ttc

act

cac

aat

caa

ag

c a

(c)1

2 12

1–16

6 58

I

41%

stm

1029

-r

aag

att

tcc

aag

aaa

ttt

gag

gg

stm

1041

-f

B

gtt

gag

tag

aag

gag

gat

t

(gaa

)5

86–1

01

53

V

13%

stm

1041

-r

cct

ttg

tct

tct

gct

ttt

g

stm

1049

-f

G

cta

cca

gtt

tg

t tg

a tt

g t

gg

tg

(a

ta)6

17

6–21

2 58

I

27%

stm

1049

-r

agg

gac

ttt

aat

ttg

ttg

gac

g

stm

2022

-f

B

gcg

tca

gcg

att

tca

gta

cta

(c

aa)3

…(c

aa)3

16

7–26

6 57

II

16%

stm

2022

-r

ttc

agt

caa

ctc

ctg

ttg

cg

stm

3009

-f

B

tca

gct

gaa

cg

a cc

a ct

g t

tc

(tc)

13

131–

211

63

VII

4%

stm

3009

-r

gat

ttc

acc

aag

cat

gg

a ag

t c

TA

BL

E3

Prim

er p

airs

use

d in

th

e p

rese

nt

stu

dy.

Tag

co

des

ref

er t

o f

luo

resc

ent

dye

s: B

= b

lue

(FA

M),

G =

gre

en (

TETR

A),

Y =

yel

low

(H

EX).

Prim

er s

equ

ence

s,

targ

et r

epea

t, a

nd

lin

kag

e g

rou

p f

rom

So

lan

um

tu

ber

osu

mfr

om

Milb

ou

rne

et a

l. (1

998)

.

213

214


RESULTSMORPHOLOGYThirty-five of the 56 characters were significantlydifferent among all taxa (Table 2) as determinedby the Tukey-Kramer test and the LikelihoodRatio Chi square test and Pearson Chi square testfor qualitative characters in JMP v. 3.1.5 (SASInstitute Inc., 1998) statistical software. Thecophenetic correlation coefficient determinedthat the UPGMA tree constructed with EUCLIDhad the highest (best) value of r = 0.835, followedby DIST (r = 0.827), MANHAT (r = 0.812), andCORR (r = 0.747). These four similarity matriceswith NJ always provided lower r scores. Theresulting UPGMA dendrogram (Fig. 2) supportedmost of the species. We will concentrate discus-sion of the primitive Mexican diploid speciesbecause this morphological study was designedmainly to search for phenetic support of these.Figure 2 is labeled as clusters A–I for ease of discussion.

Cluster A is well defined and includes all speciesand subspecies in Solanum ser. Bulbocastana(Solanum bulbocastanum subsp. bulbocastanum,subsp. dolichophyllum, subsp. partitum, and S. clarum). Both of these species perfectly supportHawkes’s (1990) concept of this series. However,Solanum bulbocastanum subsp. dolichophyllum isintermixed with subspecies partitum. Cluster Bincludes both species in Hawkes’s seriesPolyadenia (S. lesteri, S. polyadenium), likewiseproviding support for this series.

Clusters C–I include all members of Hawkes’s(1990) series Pinnatisecta, plus other 2x(1EBN)species from South America in series Circaeifolia,Commersoniana, Lignicaulia, and S. santolallae(2x[2EBN], ser. Conicibaccata). Cluster C includes S.cardiophyllum subsp. cardiophyllum and S. ×sam-bucinum (both Solanum ser. Pinnatisecta) and S.circaeifolium (ser. Circaeifolia). Solanum ×sam-bucinum was of putative origin with none of thesebut with S. pinnatisectum and S. cardiophyllumsubsp. ehrenbergii and S. pinnatisectum. Cluster Dincludes S. pinnatisectum and S. lignicaule(ser. Lignicaulia). Cluster E includes species-specificsubclusters S. tarnii and S. trifidum that group

together, the putative hybrid S. ×michoacanum(S. bulbocastanum × S. pinnatisectum), one “mis-placed” or misidentified accession of S. brachisto-trichium, the sole accession of S. hintonii (ser.Pinnatisecta), and S. commersonii (ser. Commer-soniana). Cluster F includes both accessions of S.nayaritense (ser. Pinnatisecta) and the putativehybrid S. bulbocastanum × S. cardiophyllum (con-taining members of both ser. Bulbocastana andPinnatisecta; Rodríguez et al., 1995). Cluster G con-tains all accessions of S. jamesii (ser. Pinnatisecta),and cluster H all accessions of S. cardiophyllumsubsp. ehrenbergii plus one accession of S. brachis-totrichium (ser. Pinnatisecta). Cluster I containsfour of the six accessions of S. brachistotrichiumand both replicate groups of the sole accession of S. stenophyllidium (ser. Pinnatisecta), but they donot group adjacent to each other.

In summary, the primitive Mexican diploid speciesare well supported phenetically except for S.brachistotrichium. Solanum stenophyllidium isproblematic, perhaps due to a single accessionexamined, but groups with most accessions of S.brachistotrichium. The putative hybrid S. ×sam-bucinum groups between its parents, and theputative hybrids S. ×michoacanum and S. bulbo-castanum × S. cardiophyllum away from their par-ents. The principal component analysis (notshown) presents a very similar clustering, or lackof it, regarding S. brachistotrichium.

Stepwise discriminant analysis separated all taxaby the following 21 morphological characters (cf.see Fig. 3), arranged from highest to lowest dis-criminant ability: (1) lateral leaflet decurrency(character 30, Table 2); (2) stem pubescence type(6); (3) spicy odor of leaves (34); (4) presence ofinterstitial leaflets (12); (5) ratio: length fromwidest point of most distal lateral leaflet blade tobase of blade/length of most distal lateral leafletblade (21); (6) shape of pseudostipular leaflet(33); (7) abaxial leaf pubescence length (28); (8)ratio: length of second most distal lateral leafletblade/width of second most distal lateral leaflet(24); (9) presence of stem wings (4); (10) petiolulelength of most distal lateral leaflet (22); (11)widest leaflet pair (11); (12) ratio: length of most


215

Figure 2. UPGMA dendrogram (Euclidean similarity option) based on 35 of 56 morphological characters that differ significantlyamong taxa. Species names and accession numbers, as in Table 1 and Figures 1 and 4, and phenetic groups as discussed in the text.

216


distal lateral leaflet blade/width of most distal lat-eral leaflet (19); (13) terminal leaflet width (13);(14) petiolule length of second most distal lateralleaflet (27); (15) petiole wings (31); (16) numberof leaflet pairs (9); (17) petiolule length of termi-nal leaflet (17); (18) most distal lateral leafletwidth (18); (19) length from widest point of sec-ond most distal lateral leaflet blade to base ofblade (25); (20) length from widest point of theterminal leaflet blade to base of blade (15); (21)leaf width (7).

We also used stepwise discriminant analysis todiscover characters that best distinguished thenine phenetically most similar primitive Mexicandiploid species. This analysis used the 14 morpho-logical characters, arranged from highest to low-est discriminant ability. Nine of these 14 are quan-titative character distributions and are graphical-ly displayed in Figure 3, while 5 character statedistributions are qualitative: (1) fruit shape: glo-bose in all species except S. hintonii and S. tri-fidum, which had conical fruits (56); (2) leafdecurrence: all are laterally decurrent except S.cardiophyllum subsp. ehrenbergii, S. hintonii, andS. tarnii, which are not (30); (3) shape of pseu-dostipular leaf: lunate in all species except pinnu-late in S. jamesii (33); (4) only S. ×michoacanumand S. tarnii lack interstitial leaflets (12); (5) ratio:length of second most distal lateral leafletblade/width of second most distal lateral leaflet(24); (6) abaxial leaf pubescence length (28); (7)length of calyx acumen (42); (8) petiole length(32); (9) width of second most distal lateral leaflet(23); (10) secondary color of abaxial surface ofcorolla: 100% white in S. hintonii, S. stenophyllid-ium, S. tarnii, and S. trifidum, only 42.8% white inS. brachistotrichium (the rest white with tones ofviolet), 16.6% white in S. cardiophyllum subsp.ehrenbergii and S. jamesii (the rest white withtones of violet), and 50% white in S. nayaritense(the rest white with tones of violet) (54); (11) styleexsertion (46); (12) length from widest point ofsecond most distal lateral leaflet blade to base ofblade (25); (13) stem diameter (3); (14) number ofleaflet pairs (9).

Significant differences in character states for thisreduced matrix of nine phenetically most similarprimitive Mexican diploid taxa, as determined bythe Tukey-Kramer test, found 21 characters (allquantitative) best distinguish these taxa. These 21characters include all 9 of the quantitative char-acters mentioned above as determined bySTEPDISC. The means, ranges, and standard devi-ations of these 21 characters are graphically pre-sented in Figure 3. It shows that most of thesecharacters overlap considerably in range, andprovide few species-specific character states use-ful for easy diagnoses of taxa. Some exceptions,however, are low number of lateral leaflet pairsin S. nayaritense (character 9), long petiolules(17), long petioles (32), and wide corolla lobes asmeasured at the base of the corolla junction (51)in S. hintonii. This trend of the sole to predomi-nant presence of overlapping character states todefine taxa is common in Solanum sect. Petota(Spooner & van den Berg, 1992; Spooner &Hijmans, 2001), and termed polythetic support. Apolythetic morphological species concept is onewhere species are defined where they have thegreatest number of shared features, no singlefeature of which is essential for group member-ship or is sufficient to make an organism a mem-ber of a group (Sokal & Sneath, 1963; Stuessy,1990). Stated otherwise, species are distinguishedonly by a complex set of character states thatoverlap in ranges.

MICROSATELLITE PHENETICSOnly 12 of the 45 primers we tried, from 8 linkagegroups as mapped by Milbourne et al. (1998) forSolanum tuberosum, gave useable results. Therewere amplification failures in these 12 rangingfrom 4% in stm0003 and stm3009, to 58% instm0007, with the average across all primers 22%(Table 3). It is frequent in cross-species amplifica-tion to reduce the annealing temperature inorder to obtain amplification (Ezenwa et al.,1998). A range of up to 5ºC less was attempted toachieve amplification when no product otherwisecould be obtained, but with futile results. Weconsidered an amplification successful when thepositive control (S. tuberosum) amplified, as wellas most of the primitive Mexican diploids.


217

Figure 3. Means (dot), ranges (length of line), and one standard deviation of the mean (length of box) for 21 of the 35 statistical-ly significant morphological characters of the phenetically most similar diploid wild species from the United States, Mexico, andCentral America. Included here are only the quantitative characters; the number preceding the character corresponds to Table 2.All measurements are in cm except characters 1, 3, 28, 38, 42, 46, 48, and 51, which are in mm. Three-letter species codes fol-low Spooner and Hijmans (2001): bst = S. brachistotrichium; ehr = S. cardiophyllum subsp. ehrenbergii; hnt = S. hintonii; jam = S.jamesii; mch = S. ×michoacanum; nyr = S. nayaritense; stp = S. stenophyllidium; trf = S. trifidum; trn = S. tarnii.

218


Figure 4. Neighbor-joining dendrogram based on microsatellite data (Nei & Li, 1979, similarity). Species names and accessionnumbers as in Table 1 and Figures 1 and 2, and phenetic groups from Figure 2.


219

The trees obtained from the two stepwise muta-tion model (SMM) matrices showed little to noconcordance with the morphology phenogram,that is, they failed to cluster well-defined mor-phological species (trees not presented). The infi-nite allele model (IAM) with UPGMA and NJ clus-tered taxa somewhat better but still very poorlyrelative to the morphology phenogram (Fig. 2).Our first analysis included all taxa, and NJ clus-tered taxa best (shown in Fig. 4), but still only pro-vided weak clustering of some accessions of somespecies (such as a partial clustering of Solanumcardiophyllum subsp. cardiophyllum, S. jamesii,and S. nayaritense, all members of ser.Pinnatisecta), but with failure to completely clus-ter these and most other species. To attempt bet-ter clustering of putatively related taxa, we ana-lyzed just the most phenetically similar primitiveMexican diploids S. brachistotrichium, S. cardio-phyllum subsp. cardiophyllum, subspecies ehren-bergii, S. nayaritense, and S. stenophyllidium (allwithin ser. Pinnatisecta). This reduced analysis(not presented) still intermixed accessions of allspecies except for S. cardiophyllum subsp. cardio-phyllum as in Figure 4; that is, it still showed littleto no species-specific clustering. Microsatellitesalso were of no use to investigate hybrid originsof S. ×michoacanum, S. ×sambucinum, and theputative hybrid S. bulbocastanum × S. cardiophyl-lum (Rodríguez & Vargas-P., 1994). There were noadditive species-specific bands present to testthese hypotheses.

DISCUSSIONCONCORDANCE OF MORPHOLOGICAL DATAWITH PRIOR HYPOTHESES OF RELATIONSHIPSAND CHLOROPLAST DNA DATAOur morphological results show areas of concor-dance and discordance with cpDNA results ofSpooner and Sytsma (1992), Rodríguez andSpooner (1997), and with prior hypotheses ofinterspecific relationships (Hawkes, 1989, 1990;Hawkes et al., 1988; Hawkes & Jackson, 1992). Forexample, all members of Solanum ser.Bulbocastana are well resolved and separatedfrom the rest of the Mexican diploids (Fig. 2, clus-ter A). However, S. bulbocastanum subsp. bulbo-

castanum was the only subspecies relatively wellsupported, not S. bulbocastanum subsp.dolichophyllum and subspecies partitum. Ourmorphological data also unite S. bulbocastanumand S. clarum, supporting Hawkes (1990) whounited them in Solanum ser. Bulbocastana.Chloroplast DNA data (Spooner & Sytsma, 1992;Rodríguez & Spooner, 1997), however, supportedthe sister-taxon relationship of S. bulbocastanumand S. cardiophyllum (Fig. 2, clusters A and C, inpart), not S. bulbocastanum and S. clarum.

Solanum lesteri and S. polyadenium (cluster B)form a good phenetic group, concordant withcpDNA results (Spooner & Sytsma, 1992) andHawkes’s (1990) placement in Solanum ser.Polyadenia. Solanum ×sambucinum clusters withneither of its putative parents (S. cardiophyllumsubsp. ehrenbergii and S. pinnatisectum).Solanum trifidum and S. tarnii (both ser.Pinnatisecta) also form a good phenetic group(cluster E), concordant with cpDNA data, andrecognition of their phenetic similarity by Hawkeset al. (1988). Although not examined for cpDNA,Hawkes (1990) intuitively noted a morphologicalsimilarity of S. hintonii and S. trifidum, concor-dant with our data (cluster E).

The cluster of Solanum nayaritense with theputative hybrid of S. bulbocastanum × S. cardio-phyllum (cluster F) is discordant with morpholog-ical interpretations (Rodríguez et al., 1995) as S.nayaritense has not been suggested as its pro-genitor. The phenetic relationship of S. jamesii(cluster G) and S. cardiophyllum subsp. ehren-bergii (cluster H) has never been suggestedbefore, but is reasonable based on cpDNA resultsthat show them to be part of the same clade(Spooner & Sytsma, 1992; Rodríguez & Spooner,1997). The phenetic separation of S. cardiophyl-lum subsp. ehrenbergii (cluster H) from sub-species cardiophyllum (cluster C) is concordantwith cpDNA data (Rodríguez & Spooner, 1997).The phenetic similarity of S. brachistotrichiumand S. stenophyllidium (cluster I) also has supportwith cpDNA data (Spooner & Sytsma, 1992),which places them on the same clade.

220


MICROSATELLITE DATA ANALYSISThere are two divergent hypotheses ofmicrosatellite evolution and its reflection on bestanalytical methods for assessment of genetic dis-tances and phylogeny. Schlötterer and Tautz(1992) have shown slippage during replication,which would support the use of the simple muta-tion model (SMM) for analyses. However, Colsonand Goldstein (1999) studied microsatellite loci inDrosophila and found that many mutation eventsin microsatellites are more complex than assumedby SMM. They advised the use of sequence datato assess the accumulation of imperfections with-in the microsatellite repeat, reducing the numberof consecutive repeats. These imperfections canbe insertions, deletions, or base substitutions,which, along with stepwise mutation drift tosmaller sizes, have contributed to degradation ofmicrosatellite loci over time. The authors cautionagainst the use of microsatellite size alone forpopulation or phylogenetic relationships (Colson& Goldstein, 1999). The SMM has been supportedby phylogeny reconstructions in bovine species(MacHugh et al., 1997), fish (Ruzzante, 1998), andhuman populations (Valdes et al., 1993; Shriver etal., 1995; Kimmel et al., 1996). On the other hand,the infinite allele model (IAM) has been support-ed for data sets with numerous compoundmicrosatellites (Estoup et al., 1995), which wouldbe the case in our data set. In this study, the bestconcordance with morphological data was pro-duced with the IAM and the worst with the SMM,but both were very poor. The IAM was the mostuseful in analyzing subspecies differences in S.tuberosum (Raker & Spooner, 2002).

MICROSATELLITECROSS-SPECIES AMPLIFICATIONMicrosatellites are expensive to develop. One oftheir appealing features is their potential use inrelated groups beyond the mainly economicplants where they were developed. There are several examples of successful amplificationacross species. Microsatellites distinguished twoclosely related species in genus Sitobion(Hemiptera: Aphidoidea) (Figueroa et al., 1999).Microsatellites developed for the tropical treePithecellobium elegans were conserved in several

species of the tribe Ingeae (Dayanandan et al.,1997). Microsatellites from peach successfullyamplified fragments in apple, suggesting thatthey might be useful as well in other fruit crops(Cipriani et al., 1999). In the genus Dianthus,microsatellite loci developed from the EMBLdatabase were tested in 26 species within thegenus, and successful amplification implied thatthe species are closely related or that the primerregions are well conserved (Smulders et al., 2000).Eight of 17 microsatellite primer pairs designedfor Quercus petraea were successful in producingamplifications in the related genus Castanea, and4 of the 17 in Fagus. Twelve of these 136 amplifi-cations were cloned and sequenced and shown tobe homologous to Quercus petraea; nevertheless,the authors recognized a tendency for decreasingability to successfully amplify loci, with increasingevolutionary distance across the family(Steinkellner et al., 1997). The same relationshipbetween an increasing genetic distance anddecreasing success of microsatellite amplificationwas observed in cassava (Manihot esculentaCrantz), through tests with six Manihot wildspecies (Roa et al., 2000).

In addition to problems producing microsatellitefragments across species, there also are problemswith homology of the microsatellite fragmentthat is necessary for correct phylogenetic inter-pretations (Wendel & Doyle, 1998). For successfulamplification of homologous fragments the tar-get sequence for the primers and flankingsequence about the microsatellites must be suffi-ciently conserved (Ezenwa et al., 1998).Knowledge of these criteria is costly as it requiressequencing the PCR fragments.

In our study the primers were developed througha Genbank search of Solanum tuberosum(Milbourne et al., 1998). Chloroplast DNA data(Spooner & Sytsma, 1992) show the primitiveMexican diploids to be the most distantly relatedgroup to S. tuberosum in section Petota, andmicrosatellites performed poorly for phylogenet-ic reconstructions. Similar problems have beenfound in other groups. Sequence variation and/orvariation in the number of repeats has been


found by Peakall et al. (1998), who used 31 soy-bean microsatellite loci in wild relatives andfound that cross-species amplification was low,mainly in closely related genera. When cross-species amplification was achieved there wassequence variation in the flanking regions andwithin the repeats (non-homologous amplifica-tion). Westman and Kresovich (1998) tried to useArabidopsis primer pairs to amplify microsatel-lites in the closely related genus Brassica, butmany amplified loci did not hybridize withmicrosatellite probes, and probably did not con-tain repeats. In animals, microsatellites derivedfrom two species of wasps successfully cross-species amplified in different related genera fromthe Vespidae. However, the loci from one specieswere more conserved than those from the other,warning that it is difficult to reach generaliza-tions regarding conservation rates using datafrom single species (Ezenwa et al., 1998). In avianmicrosatellites, 73 primer sets from 16 speciesacross nine families were tested on Queleaspecies. Only 22 of these 73 pairs producedhomologous amplification (Dallimer, 1999).

The issue of cross-species amplification is com-plex. As seen above, it can be achieved acrossgenera within families, or not even occur acrossspecies within a genus. To our knowledge there isno comprehensive summary that addresses thecomparative extent of microsatellite conservationor whether the failure of cross-species amplifica-tion is due to primer design or phylogenetic dis-tance. Clisson et al. (2000) sequenced primatemicrosatellites with primers designed in humansand found that alterations in the base composi-tion of repeats, and insertions and deletions inflanking regions, are important components ofthe variation but introduced homoplasy (samesize but different composition) reducing theirphylogenetic utility.

CONCLUSIONSMost of the species examined in our study weresupported by morphological characters, exceptSolanum brachistotrichium and possibly S. steno-phyllidium. However, microsatellites provided poorto no support for species. This could be caused bynon-homology of repeat length between the

primers, or in our study by not having enough datato provide phylogenetic signal. There continue tobe unanswered questions about microsatellitesregarding mutation processes, proper analyticaltools for phylogeny, and knowledge of the extentof cross-species amplification. Our results showthat microsatellites developed from S. tuberosumhave reduced utility to establish good morpholog-ical species from the phylogenetically distantUnited States, Mexican, and Central Americandiploid wild potato species. Lara-Cabrera andSpooner (2004) addressed the validity of thesespecies and their interrelationships with AFLP data.

NOTE IN PROOFThese results, and those of Lara-Cabrera andSpooner (2004), were instrumental in revising thewild potatoes of North and Central America(Spooner et al., 2004).

ACKNOWLEDGMENTSWe thank John Bamberg and staff of the UnitedStates Potato Genebank for germplasm and local-ity data; Charles Nicolet and staff of theUniversity of Wisconsin Biotechnology Center fortechnical help; Lynn Hummel and staff at WalnutStreet Greenhouse for help in growing plants; labpartners Brian Karas, Iris Peralta, Celeste Raker,and Sarah Stephenson for technical advice; PeterCrump for statistical advice; and RobertGiannattasio for artwork. We gratefully acknowl-edge the long collaboration of Bill D’Arcy, whoshowed great interest and support to us in thisproblem and many other issues in theSolanaceae. This study was supported by CONA-CYT (Mexico) scholarship number 116742 grantedto Sabina I. Lara-Cabrera, and by the UnitedStates Department of Agriculture.

LITERATURE CITEDArens, P., P. Odinot, A. W. Van Heusden, P.

Lindhout & B. Vosman. 1995. GATA- andGACA-repeats are not evenly distributedthroughout the tomato genome. Genome38: 84–90.

Areshchenkova, T. & M. W. Ganal. 1999. Longtomato microsatellites are predominantlyassociated with centromeric regions.Genome 42: 536–544.

221

222


Bichara, M., I. Pinet, S. Schumacher & R. P. P.Fuchs. 2000. Mechanisms of dinucleotiderepeat instability in Escherichia coli.Genetics 154: 533–542.

Cipriani, G., G. Lot, W. G. Huang, M. T.Marrazzo, E. Peterlunger & R. Testolin.1999. AC/GT and AG/CT microsatelliterepeats in peach [Prunus persica (L.)Batsch]: Isolation, characterisation andcross-species amplification in Prunus.Theor. Appl. Genet. 99: 65–72.

Clisson, I., M. Lathuilliere & R. B. Crouau. 2000.Conservation and evolution of microsatel-lite loci in primate taxa. Amer. J. Primatol.50: 205–214.

Colson, I. & D. G. Goldstein. 1999. Evidence forcomplex mutations at microsatellite loci inDrosophila. Genetics 152: 617–627.

Dallimer, M. 1999. Cross-species amplificationsuccess of avian microsatellites in the red-billed quelea Quelea quelea. Molec. Ecol.8: 395–398.

Dayanandan, S., K. S. Bawa & R. Kesseli. 1997.Conservation of microsatellites amongtropical trees (Leguminosae). Amer. J. Bot.84: 1658–1663.

Estoup, A., L. Garnery, M. Solignac & J. M.Cournet. 1995. Microsatellite variation inhoneybee (Apis mellifera L.) populations:Hierarchical genetic structure and test ofthe infinite allele and stepwise mutationmodels. Genetics 140: 679–695.

Ezenwa, V. O., J. M. Peters, Y. Zhu, E. Arevalo,M. Hastings, P. Seppa, J. S. Pedersen, F.Zacchi, D. Queller & J. E. Strassmann. 1998.Ancient conservation of trinucleotidemicrosatellite loci in polistine wasps.Molec. Phylogenet. Evol. 10: 168–177.

Figueroa, C., J. C. Simon, J. F. Le Gallic & H. M.Niemeyer. 1999. Molecular markers to dif-ferentiate two morphologically closespecies of the genus Sitobion. Entomol.Exp. Appl. 92: 217–225.

Goldstein, D. B., A. Ruiz-Linares, L. L. Cavalli-Sforza & M. W. Feldman. 1995. An evalua-tion of genetic distances for use withmicrosatellite loci. Genetics 139: 463–471.

Hanneman, R. E., Jr. 1994. Assignment ofEndosperm Balance Numbers to the tuber-bearing Solanums and their close non-tuber bearing relatives. Euphytica 74:19–25.

Hawkes, J. G. 1989. Nomenclatural and taxo-nomic notes on the infrageneric taxa ofthe tuber-bearing Solanums (Solanaceae).Taxon 38: 489–492.

—. 1990. The Potato: Evolution, Biodiversityand Genetic Resources. Belhaven Press,London.

— & M. T. Jackson. 1992. Taxonomic and evo-lutionary implications of the EndospermBalance Number hypothesis in potatoes.Theor. Appl. Genet. 84: 180–185.

—, J. P. Hjerting & T. R. Tarn. 1988. Two newtuber-bearing Solanum species fromMexico: S. tarnii and S. matehuelae.Phytologia 65: 114–118.

Kardolus J. P. 1998. A Biosystematic Analysis ofSolanum acaule. Thesis, WageningenAgricultural University, Wageningen, TheNetherlands.

Kimmel, M., R. Chakraborty, D. N. Stivers & R.Deka. 1996. Dynamics of repeat polymor-phisms under a forward-backward muta-tion model: Within and between-popula-tion variability at microsatellite loci.Genetics 143: 549–555.

Kimura, M. & J. F. Crow. 1964. The number ofalleles that can be maintained in a finitepopulation. Genetics 143: 549–555.

Kostia, S., M. Ruohonen-Lehto, R. Väinölä & S.L. Varvio. 2000. Phylogenetic informationin inter-SINE and inter-SSR fingerprints ofthe Artiodactyla and evolution of the Bov-tA SINE. Heredity 84: 37–45.


Lara-Cabrera, S. & D. M. Spooner. 2004. Taxon-omy of Mexican diploid wild potato(Solanum sect. Petota) species: AFLP data.Pl. Syst. Evol. 248: 129–142.

Lin, H. & M. A. Walker. 1998. Identifying graperootstocks with simple sequence repeat(SSR) DNA markers. Amer. J. Enol. Vitic. 49:403–407.

Lorieux, M., M. N. Ndjiondjop & A. Ghesquiere.2000. A first interspecific Oryza sativa ×Oryza glaberrima microsatellite basedgenetic linkage map. Theor. Appl. Genet.100: 593–601.

MacHugh, D. E., M. D. Shriver, R. T. Loftus, P.Cunningham & D. G. Bradley. 1997.Microsatellite DNA variation and the evo-lution, domestication and phylogeogra-phy of Taurine and Zebu cattle (Bos taurusand Bos indicus). Genetics 146: 1071–1086.

Milbourne, D., R. Meyer, J. E. Bradshaw, E.Baird, N. Bonar, J. Provan, W. Powell & R.Waugh. 1997. Comparison of PCR-basedmarker systems for the analysis of geneticrelationships in cultivated potato. Molec.Breed. 3: 127–136.

—, —, A. J. Collins, L. D. Ramsay, C. Gebhardt &R. Waugh. 1998. Isolation, characteriza-tion and mapping of simple sequencerepeat loci in potato. Molec. Gen. Genet.259: 233–245.

Minch, E., A. Ruiz-Linares, D. Goldstein, M.Feldman & L. L. Cavalli-Sforza. 1997.Microsat v. 1.5d: A computer program forcalculating various statistics on microsatel-lite allele data. <http://lotka.stanford.edu/microsat.html>.

Nader, W., D. Werner & M. Wink. 1999. Geneticdiversity of scarlet macaws Ara macao inreintroduction studies for threatenedpopulations in Costa Rica. Biol.Conservation 87: 269–272.

Nei, M. & W. H. Li. 1979. Mathematical modelfor studying genetic variation in terms of

restriction endonucleases. Proc. Natl.Acad. Sci. U.S.A. 76: 5269–5273.

Peakall, R., S. Gilmore, W. Keys, M. Morgante &A. Rafalski. 1998. Cross-species amplifica-tion of soybean (Glycine max) simplesequence repeats (SSRs) within the genusand other legume genera: Implications forthe transferability of SSRs in plants. Molec.Biol. Evol. 15: 1275–1287.

Petren, K., B. R. Grant & P. R. Grant. 1999. Aphylogeny of Darwin’s finches based onmicrosatellite DNA length variation. Proc.Roy. Soc. London, Ser. B, Biol. Sci. 266:321–329.

Powell, W., G. C. Machray & J. Provan. 1996.Polymorphism revealed by simple sequencerepeats. Trends Pl. Sci. 1: 215–222.

Provan, J., W. Powell & R. Waugh. 1996a.Analysis of cultivated potato (Solanumtuberosum) using intermicrosatelliteamplification. Genome 39: 767–769.

—, — & —. 1996b. Microsatellite analysis ofrelationships within cultivated potato(Solanum tuberosum). Theor. Appl. Genet.92: 1078–1084.

Raker, C. M. & D. M. Spooner. 2002. An exami-nation of the Solanum tuberosum sub-species tuberosum and subspecies andige-na using microsatellite DNA. Crop Sci. 42:1451–1458.

Rieseberg, L. H. & S. J. Brunsfeld. 1992.Molecular evidence and plant introgres-sion. Pp. 151–176 in P. S. Soltis, D. E. Soltis& J. J. Doyle (editors), Molecular System-atics of Plants. Chapman & Hall, New York.

Roa, A. C., P. Chavarriaga-Aguirre, M. C.Duque, M. M. Maya, M. W. Bonierbale, C.Iglesias & J. Tohme. 2000. Cross-speciesamplification of cassava (Manihot esculen-ta) (Euphorbiaceae) microsatellites: Allelicpolymorphism and degree of relationship.Amer. J. Bot. 87: 1647–1655.

223

224


Rodríguez, A. & D. M. Spooner. 1997. Chloro-plast DNA analysis of Solanum bulbocas-tanum and S. cardiophyllum subsp. ehren-bergii (sect. Petota). Syst. Bot. 22: 31–34.

— & O. Vargas-P. 1994. Las especies de papasilvestre (Solanum L. sección PetotaDumortier) en Jalisco. Bull. Inst. Bot, IBUG2: 1–2.

—, —, E. Villegas & D. M. Spooner. 1995. Wildpotato (Solanum sect. Petota) germplasmcollecting expedition to Mexico in 1993,with special reference to Solanum bulbo-castanum Dunal and S. cardiophyllumLindley. Potato Res. 38: 47–52.

Rohlf, F. J. 1992. NTSYS-pc v.1.70, NumericalTaxonomy and Multivariate AnalysisSystem. Applied Biostatistics, New York.

Rubinsztein, D., B. Amos & G. Cooper. 1999.Microsatellite and trinucleotide-repeatevolution: Evidence for mutational biasand different rates of evolution in differ-ent lineages. Philos. Trans., Ser. B. 354:1095–1099.

Russell, J., J. Fuller, G. Young, B. Thomas, G.Taramino, M. Macaulay, R. Waugh & W.Powell. 1997. Discriminating between bar-ley genotypes using microsatellite mark-ers. Genome 40: 442–450.

Ruzzante, D. E. 1998. A comparison of severalmeasures of genetic distance and popula-tion structure with microsatellite data:Bias and sampling variance. Canad. J. Fish.Aquatic Sci. 55: 1–14.

Saitou, N. & M. Nei. 1987. The neighbor-joiningmethod: A new method for reconstructingphylogenetic trees. Molec. Biol. Evol. 4:406–425.

SAS Institute Inc. 1998. SAS ProprietarySoftware Version 7. SAS Institute, Carry,North Carolina.

Schlötterer, C. & D. Tautz. 1992. Slippage syn-thesis of simple sequence DNA. Nucl. AcidsRes. 20: 211–215.

Schneider, K. & D. S. Douches. 1997. Assessmentof PCR-based simple sequence repeats to fingerprint North American potato cultivars. Amer. Potato J. 74: 149–160.

Shriver, M. D., L. Jin, E. Boerwinkle, R. Deka, R. E.Ferrell & R. Chakraborty. 1995. A novel genet-ic distance for highly polymorphic tandemrepeat loci. Molec. Biol. Evol. 12: 914– 920.

Smulders, M. J. M., G. Bredemeijer, W. Rus-Kortekaas, P. Arens & B. Vosman. 1997. Useof short microsatellites from databasesequences to generate polymorphismamong Lycopersicon esculentum cultivarsand accessions of other Lycopersiconspecies. Theor. Appl. Genet. 94: 264–272.

—, W. Rus-Kortekaas & B. Vosman. 2000.Microsatellite markers useful throughoutthe genus Dianthus. Genome 43: 208–210.

Sokal, R. R. & P. H. Sneath. 1963. Principles ofNumerical Taxonomy. Freeman, SanFrancisco, California.

Spooner, D. M. & T. R. Castillo. 1997.Reexamination of series relationships ofSouth American wild potatoes(Solanaceae: Solanum sect. Petota):Evidence from chloroplast DNA restrictionsite variation. Amer. J. Bot. 84: 671–685.

— & R. J. Hijmans. 2001. Potato systematics andgermplasm collecting, 1989 –2000. Amer. J.Potato Res. 78: 237–268; 395.

— & K. J. Sytsma. 1992. Reexamination of theseries relationships of Mexican andCentral American wild potatoes (Solanumsect. Petota): Evidence from chloroplastDNA restriction site variation. Syst. Bot. 17:432–448.

—, — & E. Conti. 1991. Chloroplast DNA evi-dence for genome differentiation in wildpotatoes (Solanum sect. Petota:Solanaceae). Amer. J. Bot. 78: 1354–1366.

— & R. G. Van Den Berg. 1992. Species limits andhypotheses of hybridization of Solanumberthaultii Hawkes and S. tarijense


Hawkes: Morphological data. Taxon 41:685–700.

— , R.G.Van Den Berg, A.Rodríguez, J.Bamberg,R. J. Hijmans & S. I. Lara-Cabrera. 2004. WildPotatoes (Solanum section Petota) of Northand Central America. Syst. Bot.Monogr. 68:1–209 + 9 pl.

Steinkellner, H., C. Lexer, E. Turetschek & J.Glössl. 1997. Conservation of (GA)nmicrosatellite loci between Quercusspecies. Molec. Ecol. 6: 1189–1194.

Stuessy, T. F. 1990. Plant Taxonomy: TheSystematic Evaluation of ComparativeData. Columbia Univ. Press, New York.

Swofford, D. L. 1998. PAUP*, PhylogeneticAnalysis Using Parsimony (*and othermethods). Version 4.0b3a. Sinauer,Sunderland, Massachusetts.

Tautz, D. 1989. Hypervariability of simplesequences as a general source for poly-morphic DNA markers. Nucl. Acids Res. 17:6463–6471.

Teulat, B., C. Aldam, R. Trehin, P. Lebrun, J. H.A. Barker, G. M. Arnold, A. Karp, L.Baudouin & F. Rognon. 2000. An analysisof genetic diversity in coconut (Cocosnucifera) populations from across the geo-graphic range using sequence-taggedmicrosatellites (SSRs) and AFLPs. Theor.Appl. Genet. 100: 764–771.

Valdes, M., M. Slatkin & N. B. Freimer. 1993.Allele frequencies at microsatellite loci:The stepwise mutation model revisited.Genetics 133: 737–749.

Veilleux, R. E., L. Y. Shen & M. M. Paz. 1995.Analysis of the genetic composition ofanther-derived potato by randomly ampli-fied polymorphic DNA and simplesequence repeats. Genome 38: 1153–1162.

Wendel, J. F & J. J. Doyle. 1998. Phylogeneticincongruence: Window into genome his-tory and molecular evolution. Pp. 265–296in D. E. Soltis, P. S. Soltis & J. J. Doyle (edi-tors), Molecular Systematics of Plants II:DNA Sequencing. Kluwer Academic,Norwell, Massachusetts.

Westman, A. L. & S. Kresovich. 1998. Thepotential for cross-taxa simple-sequencerepeat (SSR) amplification betweenArabidopsis thaliana L. and crop brassicas.Theor. Appl. Genet. 96: 272–281.

Winter, P., T. Pfaff, S. M. Udupa, B. Huettel, P. C.Sharma, S. Sahi, R. Arreguin-Espinoza, F.Weigand, F. J. Muehlbauer & G. Kahl.1999. Characterization and mapping ofsequence-tagged microsatellite sites in thechickpea (Cicer arietinum L.) genome.Molec. Gen. Genet. 262: 90–101.

225

taxonomy of mexican diploid wild potatoes (solanum … cabrera and spooner... · 2005. 6. 13. ·...

Documents