Biodiversity and Conservation 12: 2091–2106, 2003. 2003 Kluwer Academic Publishers. Printed in the Netherlands.

Analysis of threatened endemic trees of the WesternGhats of India sheds new light on the Red Data Bookof Indian Plants

1, 1,2*JEAN-PHILIPPE PUYRAVAUD , PRIYA DAVIDAR , JEAN-PIERRE3 4PASCAL and B.R. RAMESH

1 2Smithsonian Tropical Research Institute, P.O. Box 2072, Balboa, Panama; Salim Ali School of Ecology3and Environmental Sciences, Pondicherry University, Pondicherry 605 014, India; UMR CNRS 5558,

´Universite Claude Bernard – Lyon 1, 43, Boulevard du 11 novembre 1918, F-69622 Villeurbanne4Cedex, France; French Institute of Pondicherry, 11 St. Louis Street, PB 33, Pondicherry 605 001, India;

*Author for correspondence (e-mail: puyravaudj@tivoli.si.edu)

Received 11 February 2002; accepted in revised form 4 November 2002

Key words: Endemic species, India, Rainforest, Red Data Book, Semi-evergreen rainforest, Threatenedspecies

Abstract. The Red Data Book of Indian Plants (RDB) is a reference manual that lists threatenedangiosperms, gymnosperms and pterydophytes. Because it is widely used as an analytical tool and is amajor reference for impact assessments on vegetation, it is important that the RDB be as comprehensiveand up to date as possible. This study is an attempt to cross-check the listings in the RDB using vegetationinventories and another reference manual, the Atlas of Endemics of the Western Ghats, India [Rameshand Pascal 1997. Atlas of Endemics of the Western Ghats (India). French Institute of Pondicherry, India].Field inventories across the Western Ghats gave an estimate of the species relative abundances and theAtlas provided quantitative information on the number of records for the endemic species in herbaria,literature, and during field surveys. The results of this analysis indicate that the RDB and the Atlas agreestatistically regarding the conservation status of endemic trees. However, the proportion of threatenedspecies per Atlas record category behaves erratically, indicating that some threatened endemic trees arenot listed in the RDB. Our results suggest that the status of threatened plant species should be reexaminedon a priority basis using quantitative methodology. An updating of the RDB is urgently needed,particularly for the endemic plants of the Western Ghats. We provide a list of species that are likely to bethreatened, yet are not listed.

Introduction

There is an international effort to identify species that face extinction in order tomake conservation efforts more efficient. Since its creation, the International Unionfor the Conservation of Nature and Natural Resources (IUCN) has attempted to listspecies that are threatened with extinction globally. The criteria used to define thecategories of threatened species change with new information (IUCN 2001) and thelists of threatened species are accordingly upgraded. The Red List of ThreatenedPlants (Walter and Gillett 1998) is the most recent attempt to inventory thethreatened plant species of the World.

The Red Data Book of Indian Plants (RDB), published by the Botanical Survey of

2092

India (Nayar and Sastry 1987, 1988, 1990), focuses exclusively on the Indian flora,more precisely on threatened angiosperms, gymnosperms and pterydophytes. It is amajor publication for several reasons. Scientists studying rarity use it as ananalytical tool and the accuracy of their results depends on the accuracy of the RDB.Impact studies refer to the RDB to assess the menace of projects on plant species.Customs Agencies need the RDB to guide them in controlling the illegal trade ofbiological material under CITES. Finally, conservation organizations utilize theRDB to prioritize their efforts. Due to its widespread use, the RBD has reached thestatus of a standard reference guide with important implications for research,conservation and for the economy.

The most recent edition of the RDB is based on the criteria of Lucas and Synge(1978). Since then, the criteria used to define the categories of threatened specieshave been refined considerably (IUCN 2001). It has been shown that about 10% ofall plant taxa are threatened globally (IUCN 1994), which gives an estimate of ca.1500 of the 15000 plants recorded for the Indian subcontinent. The RDB however,contains only 814 species (5.4%) (Nayar and Sastry 1987). The relatively smallproportion of threatened plants represented in the RDB is puzzling. It is important toknow whether or not it represents an accurate estimate of the proportion ofthreatened plants in the Indian flora.

Data on plant species and their distribution in the Western Ghats are increasinglyavailable, which permits us to cross-check the listings in the RDB for a particularflora. The recently published Atlas of Endemics of the Western Ghats (India) byRamesh and Pascal (1997) examines endemic evergreen and semi-evergreen foresttrees and shrubs with a girth at breast height (GBH) greater than 10 cm. It containsquantitative information on the number of records of endemic species in herbaria, inliterature, and in field surveys. The Atlas does not give information on abundance,but ecological inventories can be used to estimate the relative abundance of species.

The objective of this study was to test whether the RDB provides a comprehen-sive list of the threatened endemic rain forest tree taxa (species, varieties orsubspecies) of the Western Ghats of India. In order to accomplish this task, we (i)listed the threatened endemic trees and shrubs of the evergreen and semi-evergreenforests of the Western Ghats, on the basis of Ramesh and Pascal (1997) and theRDB; (ii) compared the Atlas data with data gathered during ecological inventories;(iii) compared the RDB categories with the Atlas’ records; and (iv) suggest a list ofpotentially threatened taxa to be considered for inclusion in the RDB.

Methods

The Atlas of Endemic Trees included 352 tree and shrub species, subspecies, andvarieties with a GBH greater than 10 cm. Three sources of data were used forcompiling the Atlas: herbaria (almost 6000 specimens checked in the majorinternational, national and regional herbaria), literature (935 records), and fieldsurveys by teams of the French Institute of Pondicherry (India), which spannedmore than 15 years (4369 entries). The Atlas records were a composite of records (i)

2093

Tab

le1.

Spec

ies

inin

crea

sing

num

ber

ofre

cord

sin

the

Atla

san

dsp

ecie

soc

curr

ing

inth

eR

edD

ata

Boo

k.

Spec

ies

Fam

ilyR

ecor

dsR

ange

Red

Dat

aC

omm

ents

inth

eR

DB

(km

)B

ook

stat

us

H.

L.F.

I.S

.

Actin

odap

hne

lana

taM

eiss

n.La

urac

eae

11

4En

dang

ered

Berb

eris

nilg

hiri

ensi

sA

hren

dtB

erbe

ridad

eae

11

4Bu

chan

ania

barb

eri

Gam

ble

Ana

card

iace

ae1

14

Enda

nger

edC

anth

ium

perg

raci

lis(B

ourd

.)G

ambl

eR

ubia

ceae

11

4C

inna

mom

umch

emun

gian

umM

ohan

and

Hen

ryLa

urac

eae

11

4C

inna

mom

umgo

aens

eK

oste

rm.

Laur

acea

e1

14

Cin

nam

omum

heyn

eanu

mN

ees

Laur

acea

e1

14

Cin

nam

omum

wal

aiva

rens

eK

oste

rm.

Laur

acea

e1

14

Cro

ton

law

ianu

sN

imm

oEu

phor

biac

eae

11

4C

ynom

etra

bedd

omei

Prai

nFa

bace

ae(C

aesa

lpin

oide

ae)

11

4D

rype

tes

trav

anco

rica

(Bou

rd.)

S.K

.Jai

nEu

phor

biac

eae

11

4Eu

geni

aar

gent

eaB

edd.

Myr

tace

ae1

14

Extin

ct?

Ficu

san

glad

eiC

ECFi

sher

Mor

acea

e1

14

Gar

cini

aim

bert

iiB

ourd

.C

lusi

acea

e1

14

Glo

chid

ion

sisp

aren

seG

ambl

eEu

phor

biac

eae

11

4H

omal

ium

jain

iiH

enry

and

Swam

in.

Flac

ourti

acea

e1

14

Hop

eaca

nare

nsis

Hol

eD

ipte

roca

rpac

eae

11

4H

opea

jaco

biC

.E.C

.Fis

ch.

Dip

tero

carp

acea

e1

14

Rar

eK

now

nfr

omty

peon

lyH

umbo

ldtia

uniju

gaB

edd.

var.

triju

gaJo

seph

and

Cha

ndra

s.Fa

bace

ae(C

aesa

lpin

oide

ae)

11

4Ile

xga

rdne

rian

aW

t.A

quifo

liace

ae1

14

Extin

ct?

Ixor

ajo

hnso

niJ.

Hk.

Rub

iace

ae1

14

Ixor

asa

ulie

rei

Gam

ble

Rub

iace

ae1

14

Lits

eani

gres

cens

Gam

ble

Laur

acea

e1

14

Mel

icop

ein

dica

Wt.

Rut

acea

e1

14

Vuln

erab

leIm

min

entd

ange

rofe

xtin

ctio

nM

icro

trop

isde

nsifl

ora

Wt.

Cel

astra

ceae

11

4N

otho

pegi

aau

reo-

fulv

aB

edd.

Ana

card

iace

ae1

14

Enda

nger

edN

otho

pegi

abe

ddom

eiG

ambl

eva

r.w

ynaa

dica

Ellis

and

Cha

ndr.

Ana

card

iace

ae1

14

Not

hope

gia

cast

anei

folia

(Rot

h)D

ing

Hou

Ana

card

iace

ae1

14

Pitto

spor

umvi

ridu

lum

Nay

ar,G

irian

dC

hand

ras

Pitto

spor

acea

e1

14

2094

Tab

le1.

(con

tinu

ed)

Spec

ies

Fam

ilyR

ecor

dsR

ange

Red

Dat

aC

omm

ents

inth

eR

DB

(km

)B

ook

stat

us

H.

L.F.

I.S

.

Psyc

hotr

iabe

ddom

eiD

eb.a

ndG

ang.

Rub

iace

ae1

14

Sche

ffler

aro

stra

ta(W

t.)H

arm

sva

r.m

icra

ntha

(Cla

rke)

Mah

esh.

Ara

liace

ae1

14

Soph

ora

wig

htii

Bak

erFa

bace

ae(F

aboi

deae

)1

14

Sym

ploc

ospu

lchr

aW

t.ss

p.co

riace

aG

opal

anan

dH

enry

Sym

ploc

acea

e1

14

Syzy

gium

bour

dillo

nii

(Gam

ble)

Rat

hak.

Myr

tace

ae1

14

Extin

ct?

Syzy

gium

cour

talle

nse

(Gam

ble)

Als

ton.

Myr

tace

ae1

14

Enda

nger

edSy

zygi

umm

icro

phyl

lum

Gam

ble

Myr

tace

ae1

14

Syzy

gium

palg

hate

nse

Gam

ble

Myr

tace

ae1

14

Extin

ct?

Syzy

gium

utili

sR

atha

k.an

dN

.C.N

air

Myr

tace

ae1

14

Syzy

gium

zeyl

anic

um(L

.)D

C.v

arel

liptic

umH

enry

,Cha

ndra

bose

and

Nai

rM

yrta

ceae

11

4Ac

tinod

aphn

eca

mpa

nula

taJ.

Hk.

var.

obtu

saG

ambl

eLa

urac

eae

11

24

Atun

ain

dica

(Bed

d.)

Kos

term

.C

hrys

obal

anac

eae

22

4El

aeoc

arpu

sbl

asco

iW

eibe

lEl

aeoc

arpa

ceae

22

4R

are

Kno

wn

from

type

only

Elae

ocar

pus

gaus

seni

iW

eibe

lEl

aeoc

arpa

ceae

22

4R

are

Kno

wn

from

type

only

Vate

ria

mac

roca

rpa

B.L

.Gup

taD

ipte

roca

rpac

eae

11

24

Vuln

erab

lePo

lyal

thia

shen

duru

nii

Bas

haan

dSa

si.

Ann

onac

eae

22

6Sy

mpl

ocos

olig

andr

aB

edd.

Sym

ploc

acea

e2

26

Lits

eabe

ddom

eiJ.

Hk.

Laur

acea

e2

28

Mem

ecyl

onsu

bram

anii

Hen

ryM

elas

tom

atac

eae

22

8Ps

eudo

gloc

hidi

onan

amal

ayan

umG

ambl

eEu

phor

biac

eae

22

8In

dete

rmin

ate

Not

colle

cted

afte

r19

12In

gacy

nom

etro

ides

Bed

d.Fa

bace

ae(M

imos

odea

e)2

212

Extin

ct?

Ixor

ala

wso

nii

Gam

ble

Rub

iace

ae2

212

Dio

spyr

ossu

lcat

aB

ourd

.Eb

enac

eae

22

16G

loch

idio

nel

liptic

umW

t.va

r.ra

lphi

iG

ambl

eEu

phor

biac

eae

22

20Sy

mpl

ocos

mon

anth

aW

t.Sy

mpl

ocac

eae

11

232

Can

thiu

mfic

iform

eJ.

Hk.

Rub

iace

ae1

12

36Ps

ycho

tria

mac

roca

rpa

J.H

k.R

ubia

ceae

11

248

Ardi

sia

blat

teri

Gam

ble

Myr

sina

ceae

22

56

2095

Pala

quiu

mra

vii

Sasi

.Sa

pota

ceae

22

64Eu

geni

adi

scife

raG

ambl

eM

yrta

ceae

11

284

Enda

nger

edSy

zygi

umst

ocks

ii(D

uthi

e)G

ambl

eM

yrta

ceae

22

84Eu

odia

lunu

-ank

enda

(Gae

rtn.)

Mer

r.va

r.tir

unel

velic

aH

enry

and

Cha

nd.

Rut

acea

e1

12

112

Euon

ymus

pani

cula

tus

Wt.

exLa

wso

nC

elas

trace

ae2

211

2Ra

pane

ast

riat

aM

ez.

Myr

sina

ceae

11

211

2Ag

laia

mal

abar

ica

Sasi

dhar

anM

elia

ceae

11

211

6Sy

zygi

umpa

ram

esw

aran

iiM

ohan

anan

dH

enry

Myr

tace

ae2

211

6K

oelo

depa

sca

lyci

num

Bed

d.Eu

phor

biac

eae

22

120

Syzy

gium

myh

endr

ae(B

edd.

exB

rand

is)G

ambl

eM

yrta

ceae

11

212

0Ps

ycho

tria

nilg

irie

nsis

Deb

.and

Gan

g.va

r.as

teph

ana

(J.H

k.)D

eb.a

ndG

ang.

Rub

iace

ae2

212

8M

emec

ylon

sisp

aren

seG

ambl

eM

elas

tom

atac

eae

22

152

Inde

term

inat

eK

now

nfr

omty

peon

lyAc

tinod

aphn

esa

licin

aM

eiss

n.La

urac

eae

22

160

Sche

ffler

ach

andr

asek

hara

nii

Ram

am.a

ndR

ajan

Ara

liace

ae1

12

160

Syzy

gium

chav

aran

(Bou

rd.)

Gam

ble

Myr

tace

ae1

12

188

Xylo

sma

latif

oliu

mJ.

Hk.

and

Thom

s.Fl

acou

rtiac

eae

11

218

8Sy

zygi

umbe

ddom

ei(D

uthi

e)C

hith

raM

yrta

ceae

11

219

2Ag

laia

indi

ca(J

.Hk.

)H

arm

sM

elia

ceae

11

225

6M

aesa

velu

tina

Mez

.M

yrsi

nace

ae1

12

312

Euon

ymus

serr

atifo

lius

Bed

d.C

elas

trace

ae1

12

360

Enda

nger

edAr

disi

aam

plex

icau

lisB

edd.

Myr

sina

ceae

22

364

Julo

styl

ispo

lyan

dra

Rav

ian

dA

nil

Mal

vace

ae2

264

0H

umbo

ldtia

uniju

gaB

edd.

var.

uniju

gaFa

bace

ae(C

aesa

lpin

oide

ae)

21

38

Enda

nger

edEu

geni

asi

ngam

patti

ana

Bed

d.M

yrta

ceae

33

32Ex

tinct

?Eu

geni

ain

dica

(Wt.)

Chi

thra

Myr

tace

ae2

13

100

Hum

bold

tiabo

urdi

lloni

iPr

ain

Faba

ceae

(Cae

salp

inoi

deae

)1

11

310

0En

dang

ered

Sym

ploc

osna

irii

Hen

ry,G

opal

anan

dSw

amin

.Sy

mpl

ocac

eae

21

311

6An

acol

osa

dens

iflor

aB

edd.

Ola

cace

ae3

314

0C

leis

tant

hus

trav

anco

rens

isJa

bl.

Euph

orbi

acea

e2

13

288

Tare

nna

mon

ospe

rma

(Wt.

and

Arn

.)R

aju

Rub

iace

ae2

13

288

Euge

nia

coto

nifo

liaJa

cq.s

sp.c

odye

nsis

(Mun

roex

Wt.)

Ash

ton

Myr

tace

ae1

11

346

0Po

ecilo

neur

onpa

ucifl

orum

Bed

d.C

lusi

acea

e2

13

800

Inde

term

inat

eK

now

nfr

omty

peon

lyBy

rsop

hyllu

mte

tran

drum

(Bed

d.)

J.H

k.ex

Bed

d.R

ubia

ceae

44

20Eu

geni

aflo

ccos

eB

edd.

Myr

tace

ae3

14

20D

ialiu

mtr

avan

cori

cum

Bou

rd.

Faba

ceae

(Cae

salp

inoi

deae

)2

24

28In

dete

rmin

ate

Kno

wn

from

type

only

Tare

nna

agum

bens

isSu

nder

ar.

Rub

iace

ae4

472

Vuln

erab

leD

ange

rof

extin

ctio

n

2096

Tab

le1.

(con

tinu

ed)

Spec

ies

Fam

ilyR

ecor

dsR

ange

Red

Dat

aC

omm

ents

inth

eR

DB

(km

)B

ook

stat

us

H.

L.F.

I.S

.

Chi

onan

thus

linoc

iero

ides

(Wt.)

Ben

net

and

Rai

z.O

leac

eae

22

411

6C

rypt

ocar

yaan

amal

laya

naG

ambl

eLa

urac

eae

31

418

0Po

lyal

thia

rufe

scen

sJ.

Hk.

and

Thom

s.A

nnon

acea

e4

418

0Vu

lner

able

Nos

tola

chm

acr

assi

folia

(Gam

ble)

Deb

.and

Lahi

riR

ubia

ceae

12

14

188

Syzy

gium

ram

a-va

rma

(Bou

rd.)

Chi

thra

Myr

tace

ae4

420

0O

roph

eath

omso

niB

edd.

Ann

onac

eae

22

420

4G

loch

idio

npa

ucifl

orum

Gam

ble

Euph

orbi

acea

e3

14

412

Cry

ptoc

arya

stoc

ksii

Mei

ssn.

Laur

acea

e4

458

4Ag

laia

litto

ralis

Miq

.M

elia

ceae

44

680

Actin

odap

hne

bour

neae

Gam

ble

Laur

acea

e4

15

16En

dang

ered

Poss

ibly

extin

ctPh

otin

iase

rrat

ifolia

(Des

f.)K

alkm

anva

r.to

men

tosa

Ros

acea

e4

15

20(G

ambl

e)V

ivek

.and

Shet

tyLi

tsea

trav

anco

rica

Gam

ble

Laur

acea

e4

15

28G

onio

thal

amus

rhyn

chan

ther

usD

unn.

Ann

onac

eae

55

34R

are

Few

trees

obse

rved

rece

ntly

Lasi

anth

usro

stra

tus

Wt.

Rub

iace

ae5

592

Dry

pete

spo

rter

i(G

ambl

e)Pa

xan

dH

offm

.Eu

phor

biac

eae

32

513

2La

sian

thus

cilia

tus

Wt.

Rub

iace

ae2

35

132

Dys

oxyl

umbe

ddom

eiK

oste

rm.

Mel

iace

ae2

21

513

6D

ysox

ylum

ficifo

rme

(Wt.)

Gam

ble

Mel

iace

ae5

513

6Ar

disi

aso

nchi

folia

Mez

.M

yrsi

nace

ae4

15

144

Sage

raea

gran

diflo

raD

unn.

Ann

onac

eae

55

172

Enda

nger

edM

adhu

cabo

urdi

lloni

i(G

ambl

e)H

.J.La

m.

Sapo

tace

ae5

524

8Ex

tinct

?H

opea

eros

a(B

edd.

)Van

Sloo

ten

Dip

tero

carp

acea

e2

21

537

6M

eter

omyr

tus

wyn

aade

nsis

(Bed

d.)

Gam

ble

Myr

tace

ae5

541

2En

dang

ered

Glo

chid

ion

tom

ento

sum

Dal

z.Eu

phor

biac

eae

32

544

4C

ynom

etra

bour

dillo

nii

Gam

ble

Faba

ceae

(Cae

salp

inoi

deae

)3

25

500

Buch

anan

iala

nceo

lata

Wt.

Ana

card

iace

ae5

16

52D

rype

tes

wig

htii

(J.H

k.)

Pax

and

Hof

fm.

Euph

orbi

acea

e6

611

6At

una

trav

anco

rica

(Bed

d.)

Kos

term

.C

hrys

obal

anac

eae

33

612

4In

dete

rmin

ate

Not

colle

cted

in10

0ye

ars

Cin

nam

omum

filip

edic

ella

tum

Kos

term

.La

urac

eae

32

16

164

Chi

onan

thus

lepr

ocar

paTh

w.v

arco

urta

llens

is(B

edd.

)K

.K.N

.Nai

rO

leac

eae

33

616

8Ps

ycho

tria

glob

icep

hala

Gam

ble

Rub

iace

ae5

16

256

Enda

nger

edD

imor

phoc

alyx

bedd

omei

(Ben

th.)

Airy

Shaw

Euph

orbi

acea

e3

36

360

Sapr

osm

afr

agan

sB

edd.

Rub

iace

ae4

26

380

2097

Glo

chid

ion

bour

dillo

nii

Gam

ble

Euph

orbi

acea

e5

16

388

Oro

phea

eryt

hroc

arpa

Bed

d.A

nnon

acea

e6

647

2Po

pow

iabe

ddom

eana

J.H

k.an

dTh

oms

Ann

onac

eae

61

724

Rar

eR

epre

sent

edby

afe

wtre

esPi

thec

olob

ium

grac

ileB

edd.

Faba

ceae

(Mim

osoi

deae

)7

716

4C

inna

mom

umpe

rrot

tetii

Mei

ssn.

Laur

acea

e4

37

172

Syzy

gium

dens

iflor

umW

all.

exW

t.an

dA

rn.

Myr

tace

ae7

719

2Ar

alia

mal

abar

ica

Bed

d.A

ralia

ceae

61

723

2G

lypt

opet

alum

law

soni

iG

ambl

eC

elas

trace

ae7

727

6Li

tsea

ligus

trin

aJ.

Hk.

Laur

acea

e6

17

312

Can

thiu

mne

ilghe

rren

seW

t.va

r.ne

ighe

rren

seR

ubia

ceae

77

324

Lito

sant

hes

capi

tula

tus

(Wt.)

Deb

.and

Gan

g.R

ubia

ceae

61

732

8Sy

zygi

umtr

avan

cori

cum

Gam

ble

Myr

tace

ae4

12

736

4En

dang

ered

Mili

usa

nila

giri

caB

edd.

Ann

onac

eae

31

21

742

0Vu

lner

able

Kno

wn

from

type

only

Cry

ptoc

arya

bedd

omei

Gam

ble

Laur

acea

e5

11

746

4Ag

laia

cana

rens

isG

ambl

eM

elia

ceae

61

757

6Sy

mpl

ocos

anam

alla

yana

Bed

d.Sy

mpl

ocac

eae

34

18

36Pa

laqu

ium

bour

dillo

niB

rand

.Sa

pota

ceae

42

28

56In

dete

rmin

ate

Obs

erve

dre

cent

lyN

oelit

sea

fisch

eri

Gam

ble

Laur

acea

e8

868

Gar

cini

aw

ight

iiT.

And

r.C

lusi

acea

e5

12

811

2M

emec

ylon

law

soni

iG

ambl

eM

elas

tom

atac

eae

52

18

220

Mal

lotu

sat

rovi

rens

J.H

k.Eu

phor

biac

eae

53

824

8Sc

heffl

era

race

mos

aH

arm

sA

ralia

ceae

71

831

2Ix

ora

mal

abar

ica

(Den

nst.)

Mab

b.R

ubia

ceae

17

838

4Sc

heffl

era

rost

rata

(Wt.)

Har

ms

var.

rost

rata

Ara

liace

ae8

861

2Ile

xm

alab

aric

aB

edd.

Aqu

ifolia

ceae

88

676

Dio

spyr

osba

rber

iR

amas

.Eb

enac

eae

81

944

Vibu

rnum

heba

nthu

mW

t.an

dA

rn.

Cap

rifol

iace

ae7

29

170

Tare

nna

nila

giri

ca(B

edd.

)R

aju

Rub

iace

ae6

39

232

Hom

aliu

mtr

avan

cori

cum

Bed

d.Fl

acou

rtiac

eae

51

39

296

Glo

chid

ion

john

ston

eiJ.

Hk.

Euph

orbi

acea

e9

934

4H

opea

raco

phlo

eaD

yer

Dip

tero

carp

acea

e5

13

944

8Ap

orus

abo

urdi

lloni

iSt

apf.

Euph

orbi

acea

e4

59

456

Mitr

epho

ragr

andi

flora

Bed

d.A

nnon

acea

e8

19

584

Myr

istic

afa

tua

Hou

tt.va

r.m

agni

fica

(Bed

d.)

Sinc

lair

Myr

istic

acea

e8

19

660

Agla

iam

aiae

Bou

rd.

Mel

iace

ae7

29

676

Sym

ploc

osm

acro

carp

aW

t.ex

Cl.

ssp.

kana

rana

(Tal

bot)

Noo

t.Sy

mpl

ocac

eae

63

979

6G

arci

nia

trav

anco

rica

Bed

d.C

lusi

acea

e6

410

48H

opea

utili

s(B

edd.

)B

ole

Dip

tero

carp

acea

e6

13

1068

Euge

nia

calc

aden

sis

Bed

d.M

yrta

ceae

46

1088

2098

Tab

le1.

(con

tinu

ed)

Spec

ies

Fam

ilyR

ecor

dsR

ange

Red

Dat

aC

omm

ents

inth

eR

DB

(km

)B

ook

stat

us

H.

L.F.

I.S

.

Euge

nia

rottl

eria

naW

t.an

dA

rn.

Myr

tace

ae3

710

136

Syzy

gium

bent

ham

ianu

m(W

t.ex

Dut

hie)

Gam

ble

Myr

tace

ae10

1033

2C

ynom

etra

trav

anco

rica

Bed

d.Fa

bace

ae(C

aesa

lpin

oide

ae)

63

110

392

Oro

phea

unifl

ora

J.H

k.an

dTh

oms.

Ann

onac

eae

82

1042

4Vu

lner

able

Elae

ocar

pus

venu

stus

Bed

d.El

aeoc

arpa

ceae

34

411

128

Vuln

erab

leH

opea

glab

raW

t.an

dA

rn.

Dip

tero

carp

acea

e9

11

1128

0C

inna

mom

umri

pari

umG

ambl

eLa

urac

eae

62

311

480

Gar

cini

aru

bro-

echi

nata

Kos

term

.C

lusi

acea

e4

812

112

Actin

odap

hne

cam

panu

lata

J.H

k.va

r.ca

mpa

nula

taLa

urac

eae

93

1212

8C

asea

ria

wyn

aden

sis

Bed

d.Fl

acou

rtiac

eae

91

11

1224

0O

chri

nauc

lea

mis

sion

is(W

all.

exG

.Don

.)R

idsd

.R

ubia

ceae

71

31

1244

0Vu

lner

able

Very

rare

inth

reat

ened

spot

sM

emec

ylon

heyn

eanu

mB

enth

.ex

Wt.

and

Arn

.M

elas

tom

atac

eae

102

1245

2C

leis

tant

hus

mal

abar

icus

M.A

rg.

Euph

orbi

acea

e4

53

1263

6Bl

achi

are

flexa

Ben

th.

Euph

orbi

acea

e10

212

760

Hyd

noca

rpus

mac

roca

rpa

(Bed

d.)W

arb.

ssp.

mac

roca

rpa

Flac

ourti

acea

e9

13

1318

4En

dang

ered

Actin

odap

hne

law

soni

iG

ambl

eLa

urac

eae

103

1326

8Vu

lner

able

Syzy

gium

mal

abar

icum

(Bed

d.)

Gam

ble

Myr

tace

ae13

1328

0Eu

onym

usan

gula

tus

Wt.

Cel

astra

ceae

112

1328

4En

dang

ered

Cas

eari

ava

rian

sB

edd.

Flac

ourti

acea

e13

1341

2Li

tsea

glab

rata

J.H

k.La

urac

eae

94

1358

0Ac

tinod

aphn

ean

gust

ifolia

(Bl.)

Nee

sLa

urac

eae

85

1379

2Ar

disi

arh

ombo

idea

Wt.

Myr

sina

ceae

103

114

324

Mem

ecyl

onde

pres

sum

Ben

th.e

xTr

iana

Mel

asto

mat

acea

e11

21

1458

8D

rype

tes

conf

ertifl

orus

(J.H

k.)

Pax

and

Hof

fm.

Euph

orbi

acea

e9

514

680

Oct

otro

pis

trav

anco

rica

Bed

d.R

ubia

ceae

114

1511

2Ps

ycho

tria

nudi

flora

Wt.

and

Arn

.R

ubia

ceae

1515

172

Gly

cosm

ism

acro

carp

aW

t.R

utac

eae

44

715

472

Vuln

erab

leSy

zygi

umoc

cide

ntal

e(B

ourd

.)G

andh

iM

yrta

ceae

114

1578

8El

aeoc

arpu

sre

curv

atus

Cor

ner

Elae

ocar

pace

ae15

116

172

Vuln

erab

leD

iosp

yros

prur

iens

Dal

z.Eb

enac

eae

103

417

380

Term

inal

iatr

avan

core

nsis

Wt.

and

Arn

.C

ombr

etac

eae

113

317

444

Ison

andr

ape

rrot

tetia

naD

C.

Sapo

tace

ae9

35

1758

0

2099

Mili

usa

wig

htia

naJ.

Hk.

and

Thom

s.A

nnon

acea

e11

33

1766

0G

loch

idio

nm

alab

aric

umB

edd.

Euph

orbi

acea

e4

310

1778

0Sy

mpl

ocos

mac

roca

rpa

Wt.

exC

l.ss

p.m

acro

carp

aSy

mpl

ocac

eae

86

418

68Ve

rnon

iatr

avan

cori

caJ.

Hk.

Ast

erac

eae

153

1813

2D

iosp

yros

atra

taK

oste

rm.

Eben

acea

e3

1518

172

Cin

nam

omum

wig

htii

Mei

ssn.

Laur

acea

e13

23

1818

0Be

ntic

kia

cond

apan

naB

erry

Are

cace

ae8

1018

200

Vuln

erab

leD

iosp

yros

hum

ilis

Bou

rd.

Eben

acea

e9

918

216

Hum

bold

tiade

cure

nsB

edd.

exO

liv.

Faba

ceae

(Cae

salp

inoi

deae

)15

318

220

Vuln

erab

leM

icro

trop

isla

tifol

iaW

t.C

elas

trace

ae16

11

1844

4G

ymna

cran

ther

aca

nari

ca(K

ing)

War

b.M

yris

ticac

eae

171

1852

4Sc

heffl

era

capi

tata

(Wt.

and

Arn

.)H

arm

sA

ralia

ceae

81

918

568

Mal

lotu

sbe

ddom

eiJ.

Hk.

Euph

orbi

acea

e15

12

1870

4Ac

tinod

aphn

eho

oker

iM

eiss

n.La

urac

eae

1818

708

Spon

dias

indi

ca(W

t.an

dA

rn.)

Airy

Shaw

and

Form

aA

naca

rdia

ceae

171

1890

0M

icro

trop

ism

icro

carp

aW

t.C

elas

trace

ae17

11

1917

6Ps

ycho

tria

anam

alla

yana

Bed

d.R

ubia

ceae

163

1922

4G

onio

thal

amus

wyn

aade

nsis

(Bed

d.)

Bed

d.A

nnon

acea

e9

1019

284

Lito

sant

hes

venu

losu

s(W

t.an

dA

rn.)

Deb

.and

Gan

g.R

ubia

ceae

172

1930

4Ph

aean

thus

mal

abar

icus

Bed

d.A

nnon

acea

e10

45

1937

2Vu

lner

able

Kno

wn

from

type

only

Sym

ploc

osw

ynad

ense

(O.K

.)N

oot.

Sym

ploc

acea

e17

11

1939

2Ag

laia

elae

agno

idea

(Jus

s.)B

enth

.var

.bou

rdill

onii

(Gam

ble)

K.K

.N.N

air

Mel

iace

ae9

1120

52Sa

pros

ma

cory

mbo

sum

Bed

d.R

ubia

ceae

105

520

56Li

tsea

wig

htia

na(N

ees)

J.H

k.va

r.to

men

tosa

Mei

ssn.

Laur

acea

e17

320

156

Syzy

gium

tam

ilnad

ensi

sR

atha

k.an

dC

hith

raM

yrta

ceae

124

420

416

Actin

odap

hne

tadu

linga

mii

Gam

ble

Laur

acea

e14

15

2061

2N

otho

pegi

ahe

ynea

na(J

.Hk.

)G

ambl

eA

naca

rdia

ceae

112

720

628

Dry

pete

sm

alab

aric

a(B

edd.

)A

irySh

awEu

phor

biac

eae

43

1421

248

Dip

tero

carp

usbo

urdi

lloni

Bra

ndis

Dip

tero

carp

acea

e12

18

2141

2Ix

ora

leuc

anth

aH

eyne

Rub

iace

ae21

2174

0Ix

ora

elon

gata

Hey

neex

G.D

on.

Rub

iace

ae19

221

828

Agla

iaex

stip

ulat

a(G

riff.)

Bal

ak.

Mel

iace

ae9

1322

228

Pter

ospe

rmum

retic

ulat

umW

t.an

dA

rn.

Ster

culia

ceae

141

722

768

Vuln

erab

leK

ingi

oden

dron

pinn

atum

(Rox

b.ex

DC

.)H

arm

sFa

bace

ae(C

aesa

lpin

oide

ae)

282

2353

604

Vuln

erab

leEl

aeoc

arpu

sm

unro

nii

(Wt.)

Mas

t.El

aeoc

arpa

ceae

441

2065

628

Vuln

erab

le

H.5

reco

rds

inhe

rbar

ia,L

.5re

cord

sin

liter

atur

e,F.

5fie

ldre

cord

s,I.5

reco

rds

inhe

rbar

iaor

liter

atur

ew

ithou

tpr

oper

indi

catio

nof

site

colle

ctio

n,an

dS

.5to

tal

num

ber

ofre

cord

sin

the

Atla

s.

2100

in the herbaria, (ii) in literature, and (iii) in herbaria and literature without properindication of site collection (Table 1).

Species ranges were assessed by measurements on the Atlas’ maps. The ever-green and semi-evergreen rainforests lie mostly on the western slope of the WesternGhats (Figure 1) in a narrow belt parallel to the coast. The oreographic effect largelyovershadows the longitudinal effect. In this study, species ranges were equated tothe latitudinal range. The distance (in kilometers) between the two most distantpoints of observation, as reported in the Atlas, parallel to the coast was used tocalculate the range of each species. When a species was observed only once, therange was arbitrarily considered to be 4 km (1 mm on the 1/4000000 maps of theAtlas). Because the spatial data have been gathered over a long period, the speciesranges may be larger than the present range.

Fifteen plant inventories were used to assess the relative abundances of thespecies. These plant inventories ranged from 88 259 to 148 509 N latitude and the plotsizes ranged from 0.09 to 30 ha. The inventories fall into two categories according tothe lower size limit for trees chosen by the authors, i.e. 10 cm GBH or 10 cmdiameter at breast height (DBH) (Table 2). Because relative abundances of specieschange with different size limits, comparisons with the Atlas data were done forboth categories of inventories independently. The total number of field records wasobtained by summing records for each species for each lower size category, over allthe inventories. We then compared the number of records in herbaria with the

Figure 1. The location of the Western Ghats of India.

2101

Tab

le2.

Loc

atio

nan

dot

her

info

rmat

ion

pert

aini

ngto

inve

ntor

ypl

ots

used

for

the

anal

yses

.

Site

Coo

rdin

ates

Num

ber

Num

ber

Plot

Low

erR

efer

ence

sof

spec

ies

ofen

dem

icsi

zetr

eesp

ecie

s(h

a)si

ze:

10cm

Seng

alth

eri

88339

N–7

78329

E83

221

DB

HPa

rtha

sara

thy

(199

9)V

eera

paul

i88

299

N–7

78159

E74

130.

3D

BH

Swam

yet

al.

(200

0)K

akac

hi88

339

N–7

78309

E90

333.

82D

BH

Gan

esh

etal

.(1

996)

Chi

nnap

ul88

459

N–7

78179

E27

80.

09G

BH

Dav

idar

,un

publ

ishe

dK

anik

atti

88459

N–7

78159

E32

250.

09G

BH

Dav

idar

,un

publ

ishe

dK

aike

tan

98059

N–7

78169

E44

130.

09G

BH

Dav

idar

,un

publ

ishe

dV

arag

alia

r1082

59N

–768

529

E14

843

30D

BH

Ayy

appa

nan

dPa

rtha

sara

thy

(199

9)A

ttapa

di1180

09N

–7683

09E

329

0.2

GB

HPa

scal

(198

8)K

adam

akka

l1283

29N

–758

409

E56

240.

16G

BH

Pasc

al(1

988)

´U

ppan

gala

1283

09N

–758

399

E90

383.

12D

BH

Pasc

alan

dPe

lissi

er(1

996)

Nar

avi

1381

09N

–758

109

E30

170.

16G

BH

Pasc

al(1

988)

New

1382

09N

–758

109

E25

100.

14G

BH

Pasc

al(1

988)

Kog

ar1480

09N

–748

409

E11

449

2.12

GB

HG

arri

gues

(199

9)H

alag

ale

1482

19N

–758

069

E22

70.

16G

BH

Pasc

al(1

988)

Mag

od1485

19N

–748

449

E33

100.

16G

BH

Pasc

al(1

988)

2102

number of records in the inventories using a Spearman rank-order correlationcoefficient (r ), a non-parametric statistic, as the data were not normally distributeds

(Siegel and Castellan 1989). The range of the species and the number of records inherbaria were also compared using the Spearman correlation coefficient.

The mean number of records and the mean range per conservation category werecalculated. The number and proportion of threatened species was compared to thetotal number of records below the first quartile of the frequency distribution ofspecies rank.

Results

The species were ranked according to the number of Atlas records. We listed 232species with fewer than 22 records (the first quartile of the rank records) in the Atlas,as well as species mentioned in the RDB together with their threat status (Table 1).All the ‘rare’ species were changed to ‘vulnerable’ to follow the nomenclaturalrecommendations of the IUCN (1994). Species with indeterminate status or withcomments in contradiction with the threat category (reported in Table 1) wereexcluded from the calculation of the number of records per threat category of Figure 3.

Of the 352 species reported in the Atlas, 50 species (14%) belonging to 19families were listed as threatened in the RDB (Table 1). Five species withindeterminate status (not observed since the type collection) were recorded by theFrench Institute in the Atlas prior to 1997 (Ramesh and Pascal 1997). Thesespecies are: Memecylon sisparense, Miliusa nilagirica, Palaquium bourdilloni,Phaeanthus malabaricus and Poeciloneuron pauciflorum.

There was a strong correlation between the range rank and the herbaria record24rank (r 5 0.79; n 5 352; P , 10 ). In the inventories with GBH . 10 cm,s

weak but significant relationships were observed between a species’ abundanceand (i) the herbaria rank records (r 5 0.36; n 5 90; P , 0.001), and (ii) thes

species range (r 5 0.31; n 5 90; P , 0.005). Identically, in the inventories withs

DBH . 10 cm, significant relationships were observed between a species’abundance and (i) the herbaria rank records (r 5 0.29; n 5 93; P , 0.01), ands

(ii) the species range (r 5 0.32; n 5 93; P , 0.005).s

For each rank record of the Atlas, we calculated the proportion of speciesobserved in the major inventories. This proportion decreased with decreasingnumber of records in the Atlas (Figure 2). The distribution of the species

0.15x 0.15xproportions followed a logistic model ( y 5 0.193 e / (1 1 0.193 (e 2 1));2 24r 5 0.67, F 5 62.212, P , 10 ).2,30

The number of records changed significantly with increasing threat (Figure 3;F 5 7.897, P , 0.002). The Bonferroni test indicates that the category2,30

‘vulnerable’ is significantly different from the two categories ‘possibly extinct’ (P5 0.008) and ‘endangered’ (P 5 0.005). The number of threatened speciesincreased with decreasing number of records, but the proportion of threatenedspecies did not follow any particular pattern (Figure 4).

2103

Figure 2. Proportion of species observed in the pooled inventories in relation with the number ofrecords in the Atlas.

Figure 3. Number of records by threat category.

Discussion

Of the 63% endemic tree species of the evergreen rainforest and semi-evergreenforests of the Western Ghats (Ramesh et al. 1993), 14% (50 species) werethreatened according to the RDB. Of the 4500 angiosperm species (3.8% ofthreatened plants) from southern India, Daniels et al. (1995) listed 171 threatenedspecies. Matthew (1999a) reported 167 plants under stress or suspected to be

2104

Figure 4. Number of threatened species and percentage of threatened species in the different recordcategories of the Atlas; rare species only.

extinct, out of 3015 species (5.5% of threatened plant species) identified in the drypart of Tamil Nadu as well as the Palnis. A fair number of introduced plantsrightly appear in Matthew’s floras (Matthew 1983, 1999b), on which his work onthreatened species was based. Consequently, the figure of 5.5% underestimates thenumber of indigenous plants under stress. But even then, the particular assemblagestudied here, consisting exclusively of endemic species, contains a relativelyhigher proportion of threatened species than does the widespread flora of theDeccan plateau.

The Atlas records (i) correlate with the range and abundance of species in realcommunities; (ii) indicate the likelihood of finding the species in inventories; and(iii) increase with decreasing level of threat. These points indicate that the Atlasrecords and the RDB agree broadly in classifying species. However, the proportionof threatened species varies erratically within each record category, instead ofincreasing in the less recorded species categories. The lack of trend suggests thatthe RDB may have failed to detect some threatened species.

A priori, the number of records in herbaria cannot be expected to be a goodestimator of the total species abundances. Some species such as Mangifera indicaL., which are both frequent and widely distributed, are relatively poorly recorded.Field collections are not performed uniformly with the same intensity by differentbotanists. Ruokolainen et al. (2002) have suggested that widely distributed rainforest tree species are better represented in herbaria. Our results seem to supportthis suggestion. Firstly, there was a strong correlation between the range rank andthe herbaria record rank. Rare species usually have smaller ranges than abundantspecies (Gaston 1994). Similarly, herbaria records and range produced patternsanalogous to those of real communities. Secondly, there was a weak but significant

2105

correlation between the herbaria rank records and the relative abundances (as wellas ranges) of species in inventories. However, despite a significant trend, thecorrelation was too weak to infer species rank abundances from herbaria records.

The relationship between the number of Atlas records and the proportion ofthreatened species (in each record category) observed in major inventories wasbest fit by the logistic model. The probability of finding a species belonging to aparticular threat category in the field decreases with the diminishing number ofrecords in the Atlas. Conversely, when the mean number of records was plottedagainst the threat category (Figure 3) the category ‘vulnerable’ was distinct fromthe categories ‘threatened’ and ‘possibly extinct’. Because it is rarely possible toascertain whether a tree species is extinct or highly endangered, the absence ofdifference between the last two categories is expected.

These findings show that the number of records in the Atlas is a good indicatorof threat. However, the proportion of the threatened species varies erratically withthe number of Atlas records (Figure 4). It is not clear why the lowest category (1and 2 records), with 79 species, contains proportionally fewer species than the twohigher categories. Another divergence between the Atlas and the RDB can bedetected in Figure 3. The mean number of records for the category ‘vulnerable’ isapproximately 10. The cumulative frequency for 10 records included 169 specieswhen actually only 37 threatened species had been recorded. These disagreementsseem to indicate that the RDB does not record all the threatened species. Actually,the species least recorded in the Atlas have not been observed at all in fieldsurveys. For instance, none of the 39 species recorded once in the Atlas wererecorded in the field. Of these only 10 appear in the RDB. The 29 remainingspecies should probably be included in the RDB. If so, the number of threatenedtree species of the Western Ghats will significantly increase.

An urgent re-assessment of the threat categories of the least recorded species inthe Atlas of endemics is therefore highly desirable. Table 1 can be used toprioritize a revision, since it is sorted in decreasing order of suspected threat. Fivespecies, i.e. Memecylon sisparense, Miliusa nilagirica, Palaquium bourdilloni,Phaeanthus malabaricus and Poeciloneuron pauciflorum, known only from typespecimens, have been recently recorded (Ramesh and Pascal 1997) and their statusis no more indeterminate. Elaeocarpus munronii, Kingiodendron pinnatum andPterospermum reticulatum were often recorded and frequently observed in theinventories.

Acknowledgements

J.-Ph.P. thanks the Smithsonian Tropical Research Institution for support tofinalize this work. William Laurance is sincerely acknowledged for providingfacilities to J.-Ph.P. and for his useful suggestions. Annette Aiello made verypertinent editorial comments. We acknowledge an anonymous referee for hisconstructive comments.

2106

References

Ayyappan N. and Parthasarathy N. 1999. Biodiversity inventory of trees in a large-scale permanent plotof tropical evergreen forest at Varagalaiar, Anamalais, Western Ghats, India. Biodiversity andConservation 8: 1533–1554.

Daniels R.J.R., Kumar N.A. and Jayanthi M. 1995. Endemic, rare and threatened flowering plants ofSouth India. Current Science 68: 493–495.

Ganesh T., Ganesan R., Soubadra Devy M., Davidar P. and Bawa K.S. 1996. Assessment of plantbiodiversity at mid-elevation evergreen forest of Kalakad-Mundanthurai Tiger Reserve, WesternGhats, India. Current Science 71: 379–392.

´ ´Garrigues J.-P. 1999. Action anthropique sur la dynamique des formations vegetales au sud de l’Indeˆ(Ghats occidentaux, Etat du Karnataka, District de Shimoga), Ph.D. Thesis, University of Claude

Bernard, Lyon I, France.Gaston K.J. 1994. Rarity. Chapman & Hall, London.IUCN 1994. IUCN Red List Categories. International Union for the Conservation of Nature and

Natural Resources. The World Conservation Union, Cambridge, UK.IUCN 2001. IUCN Red List Categories and Criteria: Version 3.1. IUCN Species Survival Commission.

IUCN, Gland, Switzerland (Also available online at http: / /www.iucn.org / themes /ssc / redlists /RLcats2001booklet.html).

Lucas G. and Synge H. 1978. The IUCN Plant Red Data Book. IUCN, Gland, Switzerland, p. 540.Matthew K.M. 1983. The Flora of the Tamilnadu Carnatic. The Rapinat Herbarium, India (3 volumes).Matthew K.M. 1999a. A report on the conservation status of south Indian plants. Biodiversity and

Conservation 8: 779–796.Matthew K.M. 1999b. The Flora of the Palni Hills, South India. The Rapinat Herbarium, India (3

volumes).Nayar M.P. and Sastry A.R.K. (eds) 1987. Red Data Book of Indian Plants Vol. 1. Botanical Survey of

India, Calcutta, India.Nayar M.P. and Sastry A.R.K. (eds) 1988. Red Data Book of Indian Plants Vol. 2. Botanical Survey of

India, Calcutta, India.Nayar M.P. and Sastry A.R.K. (eds) 1990. Red Data Book of Indian Plants Vol. 3. Botanical Survey of

India, Calcutta, India.Parthasarathy N. 1999. Tree diversity and distribution in undisturbed and human impacted sites of

tropical wet evergreen forest in southern Western Ghats, India. Biodiversity and Conservation 8:1365–1381.

Pascal J.P. 1988. Wet Evergreen Forests of the Western Ghats of India. French Institute of Pondicherry,Pondicherry, India.

´Pascal J.P. and Pelissier R. 1996. Structure and floristic composition of a tropical evergreen forest insouth-west India. Journal of Tropical Ecology 12: 195–218.

Ramesh B.R. and Pascal J.P. 1997. Atlas of Endemics of the Western Ghats (India). Distribution ofTree Species in the Evergreen and Semi-Evergreen Forest. French Institute of Pondicherry,Pondicherry, India.

Ramesh B.R., Pascal J.P. and De Franceschi D. 1993. Distribution of endemic, arborescent evergreenspecies in the Western Ghats. Proceedings of the Rare, Endangered and Endemic Plants of theWestern Ghats. Kerala Forest Department, Kerala, India, pp. 20–29.

Ruokolainen K., Tuomisto H., Vormisto J. and Pitman N. 2002. Two biases in estimating range sizes ofAmazonian plant species. Journal of Tropical Ecology 18: 935–942.

Siegel S. and Castellan N.J. Jr. 1989. Nonparametric Statistics for the Behavioral Sciences. 2nd edn.McGraw-Hill International Editions, Statistics Series, Singapore.

Swamy P.S., Sundarapandian S.M., Chandrasekar P. and Chandrasekaran S. 2000. Plant speciesdiversity and population structure of a humid tropical forest in Tamil Nadu, India. Biodiversity andConservation 9: 1643–1669.

Walter K.S. and Gillett H.J. (eds) 1998. 1997 IUCN Red List of Threatened Plants. Compiled by theWorld Conservation Monitoring Centre. IUCN, Gland, Switzerland.