1Volution De La Mycohщtщrotrophie Chez Les Orchidщes

= == ==== ========== = = 54:7=4'9*3.7=1*=,7&)*=)*= = == = == = .8(.51.3 *=a=.414,.*=)*8=45:1&9.438=*9=(414,.*= 472&9.43=)4(947&1* =a=.414,.*=)*=1;41:9.43=*9=(414,.*= (41*=4(947&1* =a=>89T2*8=39S,7S8=*3=.414,.*`=,74342.*`= S48(.*3(*8`= >)748(.*3(*8=*9= !3;.7433*2*39= = 57S8*39S*=*9=84:9*3:*=5:'1.6:*2*39== = = 5&7= #S1&3.*=$%&= = '*=,3=2&.=,**3= = = = ;41:9.43=)*=1&=2>(4-S9S749745-.*=(-*?=1*8=47(-.)S*8= = = = = *2'7*8=):=/:7>=a= = &55479*:78=a= #_=.-.*77>=&.11*7= = 74+*88*:7`=03.;*78.9S=)*=1&=$S:3.43= = = #2*=#&7.&3,*1&= .71&3)&= 1-&7,S*=)*=$*(-*7(-*`=03.;*78.9S=)*=.:7.3`======9&1.* = = =&2.3&9*:78=a= #_=#&7(83)7S=*1488*== 74+*88*:7`=03.;*78.9S=)*=#4395*11.*7=== = = #2*= 11&.7*=5&2*8.3== 74+*88*:7`=03.;*78.9S=)*=&7.8=6= = = = #_=54>1*=#(0*>= = 74+*88*:7`=03.;*78.9S=)*=#4395*11.*7== = = #_=$.(-&7)=&9*2&3= = 74+*88*:7`=#:8S:2=) .894.7*=8&9:7*11*=)*======'43)7*8=$4>&:2*803.= = =

= = =

1 ======7S&2':1*= = = 1*99*=9-T8*=*89=&;&39=94:9=)S).S*=A=(*:==6:.=5&79&,*39=(*99*=5&88.43=(422:3*=54:7= 1*8=47(-.)S*8=*9=6:.=2439=&.)S*=94:9=&:=143,=)*=2*8=7*(-*7(-*8`=5&7=1*:7=57S8*3(*`=1*:78= (438*.18`=1*:78=6:*89.438`=1*:7=(:7.48.9S=*9=1*:7=S2*7;*.11*2*39=(422:3.(&9.+_= 03*= +&O43= )*= 7*2*7(.*7= 94:8= (*8= 47(-.)45-.1*8`= 349&22*39= 1*8= 47(-.)45-.1*8= &2&9*:78`=8*7&=)*=5:'1.*7=)*8=&79.(1*8=)*=;:1,&7.8&9.43=A=5&79.7=)*=2&=9-T8*_=<&.=)S/A=5:'1.S= 974.8=&79.(1*8=)*=;:1,&7.8&9.43`=97T8=1&7,*8=2&.8=)&:97*8=8:.;7439_=<&.=(-4.8.=)*=7S).,*7=:3*= 8>39-T8*=*3=+7&3O&.8=51:8=143,:*=*9=)S9&.11S*=6:*=3S(S88&.7*= 54:7= :3*= 9-T8*= 8:7= &79.(1*8`= &+.3=)*=+&(.1.9*7=1&=1*(9:7*=*3=+7&3O&.8=5&7=:3=51:8= 1&7,*= 5:'1.(= *9= )*= 7*3)7*= 1&= 8>39-T8*= .3)S5*3)&39*=)*=1&=1*(9:7*=)*8=&79.(1*8=8(.*39.+.6:*8_= = =

2 3 $*2*7(.*2*398= = = '&= 9-T8*= *89= ).++.(.1*2*39= &88.2.1&'1*= A= :3= ).512*= .3).;.):*1`= *9= 1*8= +S1.(.9&9.438= 6:.= 2439= S9S= &997.':S*8`= /*= 1*8= )4.8= A= 94:8= (*:== 6:.= 439= 5479S= (*= 574/*9= 4:= 6:.= 2439= S5&:1S*=5*3)&39=(*8=)*73.T7*8=&33S*8_= = = 03*=,7&3)*=5&79.*=)*=(*8=+S1.(.9&9.438`=/*=1*8=)4.8=A=#&7(83)7S=*1488*`=6:.=+29=94:7= A=94:7`=)*=574+*88*:7=A=).7*(9*:7=)*=9-T8*`=243=,:.)*=)&38=1*=243)*=)*=1&=7*(-*7(-*=*9=)*= 1*38*.,3*2*39= :3.;*78.9&.7*_= 1= 2&= (43+.S= :3= 574/*9= 5&88.433&39= *9= 1&= 1.'*79S= )*= 1*= 24)*1*7`=*9=(*89=,7B(*=A==8&=(43+.&3(*=*9=843=*89.2*=6:*=/&.=5:=2S5&34:.7=(*8=)*73.T7*8= &33S*8_=<*=3*=7*897*.38=5&8=843=.3+1:*3(*=&:=)42&.3*=)*=2&=9-T8*`=5:.86:*.1=2&=&:9&39= +&.9= (43+.&3(*= &:= 1&'47&94.7*= 6:*= )*;&39= 1*8= &25-.9-SB97*8= )S9:).&398= 4:= 1*8= 8&11*8= )*= (43,7T8_=.=/&557S(.*=&:/4:7)-:.=2&=9-T8*`=(*89=343=8*:1*2*39=54:7=1*=8:/*9=5&88.433&39= .88:=)*8=7*(-*7(-*8=)*=#&7(83)7S=6:*=54:7=1&=+472&9.43=6:*11*=&=(4389.9:S=)*=8&=5&79=*9= 1*8=1.*38=6:*11*=&=(7SS=*397*=34:8_=:94:7=)*=(*=574/*9`=#&7(83)7S=&;&.9=1:.=&:88.=9.88S=)*8= 1.*38= +4798= *9= .1= &= 8:= 2*= +&.7*= :3*= 51&(*= )&38= 843=7S8*&:= *9= 24:;7.7= A= )&:97*8= S6:.5*8`= )&:97*8= 574/*98_= 1*89= )43(= 54:7= 1*38*2'1*= )*= (*99*= .3.9.&9.43= *9= (*99*= +472&9.43= A= 1&= 7*(-*7(-*=6:*=/*=7*2*7(.*=94:9=5&79.(:1.T7*2*39=#&7(83)7S_= = = 1*5*3)&39`=8.=#&7(83)7S=&=(4397.':S=54:7='*&:(4:5=A=1&=7S:88.9*=)*=(*99*=9-T8*`= 1&= 7S&1.8&9.43= )*= (*= 574/*9= 7*548*= &:88.= 8:7= 1&.)*= *9= 1&= 5&88.43= )*8= 47(-.)45-.1*8_= <&.= )S(4:;*79=1*8=47(-.)S*8=*9=1*:78=5&88.433S8=*3=)S':9=)*=9-T8*`=*9=(*=8439=*:==6:.=5&7=1*:78= *249.438`=1*:7=(:7.48.9S`=1*:78=(433&.88&3(*8=*9=1*:78=974:;&.11*8`=2439=5*72.8=)*=7S&1.8*7= *9= (422:3.6:*7= 2*8= 7*(-*7(-*8= &;*(= 5&88.43_= 5*8= 57*2.T7*8= 7*3(4397*8= &;*(= S7&7)= <48*5-= *9= <*&3= A4*3.,= &:94:7= ) 5.54,.:2= &5->11:2 `= &:== *8(&5&)*8= 47(-.)45-.1*8= &;*(= 1&72*3= *9= .1= S9*73*1`= *37.= #&9-S`= S7&7)= $*>3&:)`= <*&38.*77*= 5*82&7&.8`= -.1.55*= *3.*?`= .1= (&55&9.((.`= <*&38-.1.55*= 3,1&)*`= B7&3(.8= 5&'433*;.11*`= 3)7S= 4:1.S`= C.3(*39= (0*7`= #.(-*1= <41>`= #.(-*1= 8.(41*= 4:= .*77*8#.(-*1= 1&.8= &:94:7= )*8= &'.348= )*= 1*5-&1&39-T7*8`= 94:9*8= +:7*39= 57S(.*:8*8= -:2&.3*2*39= *9= 8(.*39.+.6:*2*39_= #U2*= )*= 8.251*8=(4:58=)*=+.18=S9&.*39=5&7+4.8=1*=9-SB97*=)*=).8(:88.438=5&88.433S*8=&;*(=B7&3O4.8= 5:8&0`=&8(&1=*7349`=$*2>=4:(-*`=8&'*11*=141.38.4(6:&.3*=8:7=1*8=2>89T7*8=)*8=&1'.348_= !3+.3= 1*8= (43+S7*3(*8= &:57T8= )*8= &884(.&9.438= 47(-.)45-.1*8= ):= $-3*`= ):;*7,3*`= ):= '&3,:*)4(`= *9= 1478= ):= (43,7T8= )*= 1&= 84(.S9S= +7&3O&.8*= )47(-.)45-.1.*= +:7*39= )*= ,7&3)8= 242*398=)S(-&3,*8=*9=)*=7*3(4397*8`=94:/4:78=7.(-*8=)*=6:*89.438=*9=)*=34:;*&:==574/*98_= = = '4:;*79:7*=)*=348=574/*98=;*78=1&=.-&1&3)*=*9=(*99*=7.(-*=(411&'47&9.43=&;*(=&39.= F&99-&3&`=:>&33**=C*88&':97`=A&340473=7.27:&3,=3&:7&.9=5&8=;:=1*=/4:7=8&38=#&79.3*`= 74437&9= #*0./&,43,`= *9= 845541=b= 94:9= (422*= <:1.*3= $*34:19= 2&= )433S= :3= 34:;*&:= )S5&79=*3=)S3.(-&39=(*99*=+&2*:8*=89&9.43=A=&1'.348=a=,7B(*=A=;4:8`=2&=9-T8*=&=7*'43).=*9= 1*8='433*8=8:757.8*8=(422*=1*8=34:;*&:==574/*98=3439=(*88S=)S2*7,*7_= = = 1*99*=S2:18.43=8(.*39.+.6:*`=/*=)4.8=7*2*7(.*7=1*=1!B!=)*=2*=1&;4.7=+&.9=)S(4:;7.7= &:88.=&:=6:49.).*3_=03=1&'47&94.7*=41=1*8=(-*7(-*:78=8439=5&88.433S8=5&7=1*:78=7*(-*7(-*8= *9=1*:7=*3;.7433*2*39`=*9=41=1*8=5&:8*8=(&+S8=)S7.;*39=9&39=;*78=1&=)S(4:;*79*=)*=34:;*11*8= 1.'*11:1*8= 6:*= ;*78= 1&= 897:(9:7&9.43= 5->14,S3S9.6:*= )*8= (422:3&:9S8_= <*= 5&79.7&.= *3= 7*,7*99&39= 1*8= 7.7*8= *9= )S1.7*8= )*= B.33`= 1*8= (438*.18= )*= *797&3)`= 1*= 84:9.*3= )*= B7&3(0`= 1*8=

4 &3*()49*8=)*=54>1*`=*9=)*=94:8`=(*99*=).8543.'.1.9S=*9=(*99*=4:;*79:7*=6:.=)433*=&:=(*+*=(*9= *857.9=)*=1.'*79S=*9=)*=7S(7S&9.43_= = = :=1!B!`=(*89=&:88.=:3=*857.9=+&2.1.&1=6:*=/&.=)S(4:;*79=a=&:57T8=)*=#&7.*8.*77*= 6:.=&=8:=2.3.9.*7=&:==8*(7*98=)*=1&='.414,.*=241S(:1&.7*=*9=)*8=5.89*8=74:,*8=b=&:57T8=)*= #&79.3*`=6:.=2&=84:9*3:*=1478=)*=2*8=57*2.T7*8=2.88.438=b=&:57T8=)*8=&:97*8=S6:.5*8=4:= 8:7=1*=9*77&.3=)*=5S7.*3(*`=1478=)*=2&3.58=4:=)&5S7.9.+8`=41=5&794:9`=/&.=7*974:;S=(*9=*857.9= )*397&.)*=*9=)*=(43;.;.&1.9S_= = = 1*9= *857.9`= 94:8= 1*8= )4(947&398= 1*397*9.*33*39`= = *9= (*8= &33S*8= +:7*39= )*8= 51:8= *37.(-.88&39*8=9&39=):=54.39=):=;:=8(.*39.+.6:*=6:-:2&.3_=I:&97*=&33S*8=.334:'1.&'1*8`= 94:9*8=).++S7*39*8`=41=34:8=&;438=5&79&,S=348=84.7S*8`=348=+4:8=7.7*8`=348=5&88.438=a=&;*(= 33*8C.41*99*`= #&14:`= 1&74`= '4:.8*`= ':(.*= *9= *39.&3*`= 348= )S1.7*8= &79.89.6:*8= 4:= (:1.3&.7*8`= &;*(= 394.3*`= 348= ;.7S*8= (.3S2&`= &;*(= 394.3*`= #&3:`= 1*=`= .-42&8`= 348= )S,:89&9.438= )*= ;.3`= &;*(= *'`= 11&.7*`= *3`= 33*`= '*3>`= #&99-.*:`= 33&'*11*`= '.43*1= *9= 94:9*8= (*8= 84.7S*8= 2S247&'1*8`= &;*(= *3`= <:1.*3`= :.11&:2*`= >1;&.3`= 1&2.11*`= 94:9*8= (*8= 8479.*8=*9=;4>&,*8=3&9:7&1.89*8`=*9=5:.8=&;*(=1.3)>`=B7&3O4.8`=45-.*`=)*1.3*=*9=5*15-.3*= *9=94:9*8=(*8=84.7S*8=A=7*+&.7*=1*=243)*c!9=5:.8=94:/4:78`=&:==(9S8=)*=.*77*8 *37.=&;*(= 6:.=/&.=5&79&,S=94:8=(*8='*&:==242*398=)*=;.*=)*=8:757.8*=*3=8:757.8*`=*9=&;*(=6:.=/*=;&.8= 2&.39*3&39`=;*78=)&:97*8=*3(47*_= = = <*=;4:8=7*2*7(.*=94:8`=/*85T7*=8&38=9745=4:'1.*7=)*=3428`=54:7=(*8=&33S*8=*3=;497*= (425&,3.*`=A=,7&3).7=*9=A=&557*3)7*=94:8=1*8=/4:78_=<*=34:'1.*=5&8=343=51:8=(*:==6:.=8439= 1A=)*5:.8='.*3=51:8=143,9*258=a=2*8=&2.8`=)*='>43`=)*=8472&3).*=*9=)&.11*:78`=*9=5:.8=2&= +&2.11*_= #*8= 5&7*398`= 2&= 8":7`= 2439= 94:/4:78= 84:9*3:*= 5&7= 1*:7= (43+.&3(*= *9= 1*:7= *39-4:8.&82*`=2&.8=(*89=A=2*8=,7&3)8=5&7*398=6:*=/*=)4.8=51:8=*3(47*_=#43=5&54:`=*9=8*8= 7U;*8=)*8=*&:==*9=+47U98`=243=,7&3)85&5&`=*9=843=&24:7=54:7=1&=3&9:7*`=2&=2&24:=*9=8&= (:7.48.9S=.3+&9.,&'1*`=2&=,7&3)82&2&3=94:/4:78=&)2.7&9.;*_== = = !3+.3`= 2*8= 7*2*7(.*2*398= 8&((425&,3*7439= )*3(4:7&,*2*398= *3;*78= =2*8= S9:).&398==6:.=439=*++*(9:S=)*8=89&,*8=51:8=4:=24.38=5&88.433&398=&;*(=24.`=A=97.*7=)*8= ,7&.3*8`= )S(479.6:*7= )*8= 7&(.3*8= 4:= 84:5*8*7= )*8= +*:.11*8_= .4:8`= 73&:)`= 33&`= !18&`= B&33>`= 1-14S`= 1&2.11*`= B.33= *9= 1S(.1*`= ;4:8= 2&;*?= '*&:(4:5= &557.8= 5&7= ;48= 6:*89.438= 5*79.3*39*8`= ;48= .)S*8= 34:;*11*8`= ;48= 7*2&76:*8= '.*3= :9.1*8`= ;48= 5*78433&1.9S8= 94:9*8= 5&88.433S*8=*9=5&88.433&39*8`=*9=/*85T7*=6:*=;4:8=54:77*?=;4:8=S5&34:.7=94:9=&:9&39=&:= (4:78=)*=;497*=9-T8*`=6:*=)&.11*:78`=(*79&.38=(422*3(*39== = = = =

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`= 5.54,.:2=&5->11:2 =*89=&884(.S*= 85S(+.6:*2*39=A=)*8=(-&25.,3438=*(942>(47-.?.*38 ______., ,_/ 1422*=1*8=3*499.S*8=9*25S7S*8`=1*8=3*499.S*8=9745.(&1*8=8439=&884(.S*8=A=)*8= (-&25.,3438=*(942>(47-.?.*38______., ,_0 B1:==)*=(&7'43*=*9=1#8=84:8=1*8=9745.6:*8= ______.. _ &=343885S(.+.(.9S=)*8= 5->1147(-.8 ______.2 -_+ 03*=8:757.8*=)&38=1*8=7&(.3*8=)*8= 5->1147(-.8 ______.2 -_, '&=343885S(.+.(.9S=a=:3*=5&79.(:1&7.9S=9745.(&1*= ______.2 -_- 03*=5&79.(:1&7.9S=)*8=+47U98=)8.*=):=:)8*89=______/* -_. 03*=5&79.(:1&7.9S=)*8= 5->1147(-.8 = ______/. -_/ '*8=?43*8=)42'7*8=) 5->1147(-.8 ______/0 _ *=1&=2.=49745-.*=A=1&=2>(4-S9S749745-.* ______0* ._+ '&=2.=49745-.*=*9=843=+43(9.433*2*39 ______0* ._, '&55&7.9.43=*9=1*=2&.39.*3=)*8=&1'.348=)*8=3*499.S*8 ______0. ._. '*=2&.39.*3=)*=1&1'.3.82*______12 _ S;41:9.43=)*=1&=2>(4-S9S749745-.*=(-*?=1*8=47(-.)S*8=7*;.8.9S*______2, /_+ '*=2&.39.*3=)*8=2.=49745-*8 ______2, /_, 5*=1&=2.=49745-.*=A=1&=2>(4-S9S749745-.* ______2. /_- 5*=1&:949745-.*=A=1&=2.=49745-.*______22 /_. '*=7S,.2*=)*=7*574):(9.43=)*8=2>(4-S9S749745-*8 ______3, /_/ '*8=3*499.S*8=*9=1&=2>(4-S9S749745-.* ______30 43(1:8.438=*9=5*785*(9.;*8______32 ;; ______+*, =

6 .'!=5!=B 0$!= = B.,:7*=+_=97:(9:7*=)*8=).++S7*39*8=2>(47-.?*8______+/ B.,:7*=,_= 4349745&=->545.9->8`= :3*=!7.(&(S*=2>(4-S9S749745-* ______+0 B.,:7*=-_=11&88.+.(&9.43=5->14,S3S9.6:*=)*8=3,.485*72*8`=57S8*39&39=54:7=(-&6:*=(1&)*=1*=9>5*= )42.3&39=)&884(.&9.43=2>(47-.?.*33* ______+1 B.,:7*=._='*8=8>3)72*8=)*8=51&39*8=2>(4-S9S749745-*8______+3 B.,:7*=/_= &((&7.&=&2*9->89.3& ______,3 B.,:7*=0_=11&88.+.(&9.43=5->14,S3S9.6:*=)*8=84:8897.':8=)%7(-.)S*8______-+ B.,:7*=1_=$S5&79.9.43=)*8=3*499.S*8______-/ B.,:7*=2_=7'7*=5->14,S3S9.6:*=)*8=3*499.S*8______-1 B.,:7*=3_=48.9.43=5->14,S3S9.6:*=)*8=).++S7*398= 5.54,.:2=&5->11:2 =):=<&543`=$:88.*=*9=B7&3(*_ __ -2 B.,:7*=+*_=->14,S3.*=)*8=%7(-.)S*8=7S&1.8S*=A=5&79.7=):3=,T3*=(-147451&89.6:*`= 58& ______-3 B.,:7*=++_ 5.54,.:2=&5->11:2 =*9=8*8=2>(47-.?*8______.+ B.,:7*=+,_=!(414,.*=)*8=(-&25.,3438=2>(47-.?.*38=)*=5->1147(-.8=2439&3& =#`= _=(&:)&9& =1= *9= *5-&1&39-*7&=*=.,:& =1!______.- B.,:7*=+-_=(-S2&8=(425&7&9.+=):=+43(9.433*2*39=)*=1&=2>(4-S9S749745-.*=*9=)*=1&=2.=49745-.*_ ./ B.,:7*=+._=5.;*78.9S=)*8=(-&25.,3438=.)*39.+.S8=)&38=1*8=7&(.3*8=) 5->1147(-.8=2439&3& ______.1 B.,:7*=+/_=5.;*78.9S=)*8=(-&25.,3438=*(942>(47-.?.*38=)&38=1*=347)=)*)=1&=.-&1&3)*______.3 B.,:7*=+0_=-494,7&5-.*8= .3=8.9: =)*8=&55&7*.18=;S,S9&9.+8=*9=7*574):(9*:78=&S7.*38=)*= _=2439&3& =*9= )*= _=*=.,:& =.3=8.9:______/+ B.,:7*=+1_=>89T2*8=7&(.3&.7*8=) _=2439&3& =*9=)*= _=*=.,:& ______/- B.,:7*=+2_=$*57S8*39&9.43=8(-S2&9.6:*=)*=1&=7S5&79.9.43=)&38=1*85&(*=)*8=).++S7*398=(-&25.,3438= .)*39.+.S8=A=5&79.7=)*8=7&(.3*8=)*=)*:==.3).;.):8______// B.,:7*=+3_=$S5&79.9.43=):=,*37*= 5->1147(-.8 =*3=8.*______/1 B.,:7*=,*_=4:7(*39&,*8=)-S9S749745-.*=*3=(&7'43*=2*8:7S8=8:7=)*8=+*:.11*8=)*= _=)&2&843.:2 =;*798= )*=2&78=A=/:.3=,**1______/3 B.,:7*=,+_=C&7.&9.438=.39*7847,&3*8=):=54:7(*39&,*=)-S9S749745-.*______0+ B.,:7*=,,_='*8=&1'.348=)*8=3*499.S*8______0- B.,:7*=,-_=$*574):(9.43=*9=897:(9:7*=85&9.&1*=)*= _=)&2&843.:2 ______0/ B.,:7*=,._=B&(9*:78=5*72*99&39=1&55&7.9.43=*9=1*=2&.39.*3=)*8=&1'.348______00 B.,:7*=,/_=-*34,7&22*=4'9*3:=A=5&79.7=):=541>2475-.82*=,S3S9.6:*=B' ______01 B.,:7*=,0_=411.3.*8=5*3)&39*8=):3*=+1*:7=4:;*79*=)*= _=)&2&843.:2 ______01 B.,:7*=,1_=45:1&9.43= )*=*5-&1&39-*7&=)&2&843.:2 ______03 B.,:7*=,2_=842'7*=)*=9.,*8=;*79*8=4:=&1'.348=&:=(4:78=)*=1&=8&.843______1+ B.,:7*=,3_=!39424+&:3*=57S8*39*=8:7=1*8=9.,*8`=+*:.11*8=*9=+7:.98=)*= _=)&2&843.:2 =&1'.348______1- B.,:7*=-*_=B*79.1.9S=(425&7S*=)*8=.3).;.):8=&1'.348=*9=;*798=)*=1_=)&2&843.:2______1/ B.,:7*=-+_=5>3&2.6:*=)*8=)*:==545:1&9.438=)*= _=)&2&843.:2 ______11 B.,:7*=-,_=.1&3=)*8=+&(9*:78=&++*(9&39=1&=;&1*:7=8S1*(9.;*=)*8=.3).;.):8=&1'.348______13 B.,:7*=--_=55&7.9.43=*9=2&.39.*3=)*8=&1'.348=)&38=1*8=*85T(*8=2.=49745-*8______2* B.,:7*=-._=C&7.&9.438=.397&545:1&9.433*11*8=*9=.397&85S(.+.6:*8=):=(479T,*=)*=2>(47-.?*8=(-*?= _= )&2&843.:2 =*9=;&7.&9.438=.39*785S(.+.6:*8=(-*?=1*8=3*499.S*8 ______2+ B.,:7*=-/_=1425&7&.843=)*= _=)&2&843.:2 =&1'.34`=*9=)*=)*:==*85T(*8=2>(4-S9S749745-*8`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`=4.,3*;.11*=*9=#439+*77.*7______1* .&'1*&:=++_=&7&2T97*8=+41.&.7*8=2*8:7S8=8:7=1&=545:1&9.43=)*= _=)&2&843.:2 ______1, .&'1*&:=+,_=745479.43=).3).;.):8=5&7&8.9S8=4:='74:9S8 ______1. .&'1*&:=+-_=1425&7&.843=)*=1&=;&1*:7=8S1*(9.;*=+*2*11*=)*8=.3).;.):8=&1'.348=*9=)*8=.3).;.):8=;*798= A=#439+*77.*7=*9=4.,3*;.11*______10 .&'1*&:=+._=1&8=)&1'.3.82*=949&1=+*:.11*`=9.,*`=+1*:7=(-*?=)*8=*85T(*8=)47(-.)S*8______3* = == !3(&)7S=+_=#S9-4)*8=(74.8S*8).)*39.+.(&9.43=)*8=(-&25.,3438=2>(47-.?.*38=*9=)*=1&______,+ !3(&)7S=,_='&='.414,.*=)*8=47(-.)S*8=7S8:2S*=*3=/=54.39=*88*39.*18=a______,/ !3(&)7S=-_='*8=7-.?4(943.&8$______,0 !3(&)7S=._='*8=7S8*&:==*(942>(47-.?.*38=(422:38 ______,1 !3(&)7S=/_='*8=2>(4-S9S749745-*8=a=)*8=5&7&8.9*8=______-- !3(&)7S=0_='*=7*,.2*=7*574):(9*:7=)*= _=)&2&843.:2 ______01 =

8 9 = = = '&= 5-4948>39-T8*= *89= :3= 574(*88:8= 2S9&'41.6:*= 6:.= 5*72*9= )*= (43;*79.7= ):= (&7'43*= .347,&3.6:*= *3= (&7'43*= 47,&3.6:*`= 84:8= +472*= )*= 8:(7*8`= ,7B(*= A= 1S3*7,.*= 1:2.3*:8*_= .38.`= 147,&3.82*= 5-4948>39-S9.6:*= *89= :3= &:949745-*= 6:.= 8>39-S9.8*= 8&= 57457*= 2&9.T7*= 47,&3.6:*_= '*8= 57*2.T7*8= 97&(*8= )*= 5-4948>39-T8*= 7*2439*39= A= -`2= 2.11.&7)8= )&33S*8`= 574'&'1*2*39= (-*?= )*8= &3(U97*8= )*8= (>&34'&(9S7.*8= &(9:*11*8_= .= 1*8= ;S,S9&:== *:(&7>49*8= &(9:*18= 8439= (&5&'1*8= )*++*(9:*7= 1&= 5-4948>39-T8*= ,7B(*= A= 1*:78= 51&89*8`= (*89= 6:*= (*:=8(.= )S7.;*39= )*= 1*3)48>2'.48*= ):3*= (>&34'&(9S7.*= *9= ):3= 57494?4&.7*`=.39*7;*3:*=.1=>=&=*3;.743=+`/=2.11.&7)8=)&33S*8=7(-.'&1)`=,**3_=:.9*=A=(*99*= *3)48>2'.48*= .3.9.&1*`= 1*8= 51&89*8= 439= S9S= 97&38+S7S8= A= 51:8.*:78= 7*57.8*8`= 5&7= 1*= '.&.8= )*3)48>2'.48*8= 8*(43)&.7*8= 4:= 9*79.&.7*8= /479-= *9= &1_` = ,**.= 4:= 439= S9S= &:88.= +7S6:*22*39= 5*7):8`= 5&7= *=*251*= (-*?= 1*8= %42>(T9*8`= 1*8= 1.1.S8`= 1*8= .7>5&34842*8= *9= 349&22*39= (-*?= 1&824).:2= +&1(.5&7:2 `= 1&,*39= )*= 1&= 2&1&7.&= F.11.&2843= *9= &1_` = +33.b= F.1843= *9=&1_` =+330=b= A7&:8*`=,**2b=$*>*887.*94= *9=&1_`= ,**2b=&1)&77.&,&= *9=&1_`= ,**+ _== 1-*?=1*8=;S,S9&:==9*77*897*8=*2'7>45->9*8`=1&=+43(9.43=5-4948>39-S9.6:*=&=&:88.= S9S= +7S6:*22*39= 5*7):*= 8&38= 6:*= 1*= (-147451&89*= 3*= ).85&7&.88*= *39.T7*2*39= 543(9:*11*2*39=(422*=(-*?=)*8=2:9&398=)*=1&'47&94.7*=4:=)*8=(:19:7*8= .3=;.974 =*37.(-.*8= *3= 8:(7*8= B*1)2&33`= +33+=b= *>71*= ]= 2.9-`= +33- `= *3= 7S5438*= A= )*8= 897*88= *3;.7433*2*39&:== .:'&= *9= &1_` = +330=b= A:78&7`= ,**-=b= $4'*79= *9= &1_` = ,**/`= 4:= )*= +&O43= 5*72&3*39*= (-*?= (*79&.38= 9&=438= 9*18= 6:*= 1*8= %74'&3(-*8_= 1*8= )*73.T7*8= 8439= )43(= -S9S749745-*8`=*9=5&7&8.9*39=)&:97*8=51&39*8=&:949745-*8=&:=24>*3=)*=8:O4.78_=1*=24)*= )*= ;.*= 5&7&8.9*= *89= &55&7:= (-*?= += )*8= 3,.485*72*8= *9= )&38= ++= +&2.11*8= ).++S7*39*8= &702&3`= ,**1=b= 2&.8= (*= 3*= 8439= 5&8= 1A= 1*8= 8*:1*8= 51&39*8= -S9S749745-*8_= 5&:97*8= 3,.485*72*8= 343= 5-4948>39-S9.6:*8= 1*:78= 7*88*2'1*39= 8&38= 94:9*+4.8= 57S8*39*7= )*= 1.*3= 5->8.6:*= ).7*(9= &;*(= :3*= 51&39*= &:949745-*=b=*11*8= +:7*39= 143,9*258= (438.)S7S*8= (422*= )*8=51&39*8==8&5745->9*8=`=(*898A8).7*=6:.=8*=34:77.88*39=)*=1&=2&9.T7*=47,&3.6:*=2479*= ):=841_='&=7S&1.9S=*89=94:9*=&:97*`=*9=1*8=2>89T7*8=)*=(*99*=3:97.9.43=-S9S749745-*`=S1:(.)S8= 97T8=7S(*22*39`=439=5*72.8=).)*39.+.*7=343=5&8=:3*=84:7(*`=2&.8=:3=;S7.9&'1*=7S8*&:=)*= 84:7(*8`=*9=)*=(.7(:1&9.43=):=(&7'43*=47,&3.6:*=)&38=1*=841_= = := (4:78= )*= 2&= 9-T8*= 8:7= 1S;41:9.43= )*= (*= 24)*= )-S9S749745-.*= (-*?= 1*8= 47(-.)S*8`=/&.=(-*7(-S=A=7S543)7*=A=51:8.*:78=6:*89.438=a= = 8 6:*18= +&(9*:78= )S9*72.3*39= 1&884(.&9.43= )*8= 47(-.)S*8= A= 1*:7= 84:7(*= )*= (&7'43*= 47,&3.6:*= = )*8= (-&3,*2*398= )-9*8= 84398.18= 5488.'1*8=)&38= 1S;41:9.43=)*=(*8=47(-.)S*8= 8 1&= 84:7(*= )*= (&7'43*= 47,&3.6:*= *898*11*= 1&= 2U2*= )&38= ).++S7*398= S(48>89T2*8==*:9843=(425&7*7=1*8=24)T1*8=9745.(&:==*9=9*25S7S8== 8 (422*39= (*= 24)*= )*= 3:97.9.43= *898.1= &55&7:= (-*?= 1*8= 47(-.)S*8== 6:*3= *898.1=)*8=51:8=574(-*8=5&7*398=)*8=47(-.)S*8=-S9S749745-*8`=84398.18=*:== &:88.= -S9S749745-*8`= &:= 24.38= *3= 5&79.*== *:;*398.18= 34:8= &.)*7= A= (4257*3)7*=1S;41:9.43=)*=1-S9S749745-.*=(-*?=1*8=47(-.)S*8== = =

10 11 == = =

_ _= > `= #_8_= 50%`= #_= $%BB.`= '_= C81!8%.`= !_= 5!#$= ]= #_8_= !'%!_=!;.)*3(*=+742=545:1&9.43=,*3*9.(8=9-&9=9-*=*(942>(477-.?&1='&8.).42>(*9*= &((&7.&= &2*9->89.3& = .8= &3= &(9:&1= 2:19.8-489= 8>2'.439_= 41*(:1&7= (414,> `= ,**2`= ;41= +1= .88:*=+,`=55=,2,/8,2-2_= =

_ _= >` = ._= & #!`= #_= &#.%`= A_= F!`= 1_= !8Z`= _= B11%`= _= %8B8.!`= #_= :_= !'%! _= !(942>(477-.?&1= !34(>'* = 85*(.*8= &884(.&9*= <.9-= 9-*= 2>(4-*9*749745-.(= 47(-.)= 5.54,.:2= &5->11:2 = ':9= 349= <.9-= .98= &8*=:&1= 5745&,:1*8_= 33&18=4+=49&3>`= ,**3`=5:'1.8-*)=431.3*_= =

_ _= > `=_=F.. 8`=_=.!$`=B_=$1 $5`=_=C!0.$`=#_=:_=!'%!_= #>(4-*9*749745-.(= 47(-.)8= +742= .-&.1&3)= 9745.(&1= +47*898= &884(.&9*= <.9-= &= '74&)= ).;*78.9>=4+=*(942>(477-.?&1=+:3,.`=,**3`=#1=.414,>_= =

_ #_8_=!'%!=]= _= >=,**3 = 7**3=51&398=9-&9=+**)=43=+:3,.a=+&(98=&3)=6:*89.438= &'4:9=2.=49745->_= "7*3)8=.3= 1&39=#(.*3(*8 =+.,a=/380-_= =

_ C_= .$81 58'%#$5%` = _= >`= !_= 0 %.`= _= 8.%$%`= 0_= 0...!`= #_8_=!'%!`=_=1%ZZ%'8%_=5&9.&1=7*5&79.9.43=&3)=,*3*9.(=7*1&9.438-.5=4+=,7**3= &3)= 34385-4948>39-*9.(= .3).;.):&18= .3= 2.=*)= 545:1&9.438= 4+= *5-&1&39-*7& = 47(-.)8`= ,**3`=&((*59*)=.3= 1&39=.414,> _= =

_ _= > `=_=.% 8!..`=<_=$!8%0'.`=<_81_=. %#`='_=#%8`= _= 1 .Z`= 1_= 5#!8`= #_8_= !'%!_= 84385-4948>39-*9.(= &1'.34= ;&7.&398= .3= *5-&1&39-*7&= )&2&843.:2 a=&=)*&)8*3)=<&>=.3=*;41:9.43=94<&7)=2>(4-*9*749745->=3=57*5_ = =

_ _= ]_= !5!$!8`= _= F.. 8`= _= > `= _= 055!!`= #_= :_= !'%!_= *5-&1&39-*7&=*=.,:& =7*).8(4;*7*)a=3*<=.38.,-98=.3=9-*=9&=4342>`=-&'.9&9=7*6:.7*2*398= &3)= '7**).3,= 8>89*2= 4+= &= 7&7*= 2>(4-*9*749745-.(= 47(-.)`= ,**3`= &((*59*)= .3= $47).(= %4:73&1=4+=49&3>_= =

_ _= >= ]=#_=:_=!'%!_=!(414,.*=)*=6:*16:*8=47(-.)S*8=+47*89.T7*8=a=)*8=(1*+8=54:7= 1*:7=(438*7;&9.43_=(9*8=)*=1&=/4:73S*=8(.*39.+.6:*=):=5&7(=)*8= 7&3)8=1&:88*8`=#.11&:`= 8*59*2'7*=,**1_=5_=0+802_= =

12 = = -&25.,343 = %7,&3.82*=+.1&2*39*:== -S9S749745-*=A= 5&74.=(-.9.3*:8*_= '*8= (-&25.,3438= S9:).S8= &:= (4:78= )*= (*99*= 9-T8*= 8439= *88*39.*11*2*39= )*8= &8.).42>(T9*8= A= (14.8438= *9= +472&39= )*8= '&8.)*8= 1478= )*= 1&= 2S.48*= 4:= )*8= 8(42>(T9*8= A= (14.843= *9= +472&39= )*8= &86:*8_='*8= 142S742>(T9*8=8439=343=(14.8433S8`=8439=94:/4:78=8>2'.49.6:*8=)*=51&39*8= +472&39=)*8=2>(47-.?*8=;S8.(:148&7':8(:1&.7*8=*9=1*:7=7*574):(9.43=*89=2S(433:*_=

= &8.).42>(T9*8=74:.11*8`=(-&7'438`== = (-&25.,3438=A=(-&5*&:=)439=7-.?4(943.&======8(42>(T9* 8=97:++* = = 142*742>(T9*8 = = Z>,42>(T9*8 = = 1->97.).42>(T9*8 = = 3.2&:= = = >(47-.?*8 = 884(.&9.438= ):7&'1*8= *397*= 1*8= 7&(.3*8= ):3*= 51&39*= *9= 1*= 2>(S1.:2= ):3= (-&25.,343=41=1*=(-&25.,343=+4:73.9=)*=1*&:=*9=)*8=8*18=2.3S7&:==*9=7*O4.9=)*8=8:(7*8= 574):.98=5&7=1&=51&39*_= = &5745->9*=4:=8&5749745-* = I:.=8*=34:77.9=)*=2&9.T7*=47,&3.6:*=)47,&3.82*8=24798=*==a= (-&25.,3438=6:.=)S,7&)*39=1&=2&9.T7*=47,&3.6:*=):=841_= = 3)45->9* = %7,&3.82*=6:.=;.9=)&38=1*8=9.88:8=;.;&398=):3*=51&39*`=,S3S7&1*2*39=)&38= 1*8=+*:.11*8`=8&38=)S97:.7*=1&=(*11:1*_== = :9:&1.82* = $*1&9.43=):7&'1*=*397*=)*:==47,&3.82*8=)*85T(*8=).++S7*39*8=)439=1*8=)*:== 7*9.7*39= )*8= 'S3S+.(*8_= 5&38= 1*= (&8= ):3*= &884(.&9.43= ):7&'1*= )&38= 1*= 9*258= 6:.= 8*= 2&9S7.&1.8*=5&7=:3*=(4*=.89*3(*=5->8.6:*`=.1=8&,.9=):3*=8>2'.48*_= = &7&8.9.82* = $*1&9.43=):7&'1*=*397*=)*:==47,&3.82*8=)*85T(*8=).++S7*39*8=41=1:3*=)*8= )*:==*85T(*8=7*9.7*=)*8='S3S+.(*8=)*=1&884(.&9.43=2&.8=5&8=1&:97*_=1422*=1*=(429=):3*= &884(.&9.43= )S5*3)= )*8= *85T(*8= 6:.= .39*7&,.88*39= *9= )*8= (43).9.438= *=9S7.*:7*8`= 1&= 1.2.9*= *397*=8>2'.48*=*9=5&7&8.9.82*=3*89=5&8=+.=S*=*9=1*8=).++S7*39*8=&884(.&9.438=7*57S8*39*39=:3= (439.3::2=*397*=8>2'.48*=*9=5&7&8.9.82*_= = :949745-* = I:.=574):.9=8&=57457*=2&9.T7*=47,&3.6:*=A=5&79.7=)*=1&=+472*=.347,&3.6:*= ):3= S1S2*39= *=_= 1&7'43*= *9= ,7B(*= A= :3*= 84:7(*= )S3*7,.*= *== a= ;S,S9&:== 5-4948>39-S9.6:*8=6:.=574):.8*39=)*8=8:(7*8=A=5&79.7=)*= ,%=*9=1% ,=*9=,7B(*=A=1S3*7,.*= 1:2.3*:8*_= = S9S749745-* =I:.= 8*= 34:77.9= )*= 1&= 2&9.T7*= 47,&3.6:*= 574):.9*= 5&7= :3= &:97*= 47,&3.82*= *==a=(-&25.,3438`=&3.2&:=`=51&39*8=5&7&8.9*8=4:=2>(4-S9S749745-*8_= = .=49745-* = I:.= *89= A= 1&= +4.8= &:949745-*= *9= -S9S749745-*= 54:7= 1*= 2U2*= S1S2*39`= 5&7= *=*251*`=*3=:9.1.8&39=A=1&=+4.8=:3*=84:7(*=.347,&3.6:*=*9=47,&3.6:*=)*=(&7'43*_=1*=24)*= )*=3:97.9.43=*=.89*=(-*?=)*8=&1,:*8=51&3(943.6:*8`=*9=6:*16:*8=51&39*8=)*8=47(-.)S*8`=)*8= S7.(&(S*8`=c=

13 = >(4-S9S749745-* = S9S749745-*= 6:.= 8*= 34:77.9= )*= 2&9.T7*= 47,&3.6:*= +4:73.*= 5&7= )*8= (-&25.,3438_= = -.?4(943.& = '*8=7-.?4(943.&8$=8439=)*8=(-&25.,3438='&8.).42>(T9*8=6:.=&55&79.*33*39= A= )*8= ,74:5*8= ).89&398= 5->14,S3S9.6:*2*39_= 18= 439= S9S= 7*,74:5S8= 84:8= 1*= 9*72*= =7-.?4(943.&$= ):= +&.9= )*= 1*:78= 7*88*2'1&3(*8= 2475-414,.6:*8= &:= 89&)*= &8*=:S=a= *3= (:19:7*= .18= +472*39= :3= 9-&11*= '1&3(= A= 2&7743= +43(S= A= (74.88&3(*= 7&5.)*=b= 1*:7= 2>(S1.:2= 57S8*39*=)*=+7S6:*39*8=(43897.(9.438=&:=3.;*&:=)*8=8*59&=(14.8438`=)*8=7&2.+.(&9.438=&;*(= :3*=47.*39&9.43=)*=./J=A=3*J_== = = = ;; = = %= 45:1&9.43=)*= _=)&2&843.:2 =)*=4.,3*;.11*= = 1= 1&7'43*= = 1#8= $S8*&:=#>(S1.*3=1422:3= = !1#= 1-&25.,343=*(942>(47-.?.*3= = .= !85&(*:7=39*73*=.7&38(7.9=)*=158=7.'4842&1=3:(1S&.7*_= = # = #>(4-S9S749745-*= = #%8= 45:1&9.43=)*= _=)&2&843.:2 =)*=#439+*77.*788:78'*?= = 8= ?49*= = = &7&8.9*= = $= $&).&9.43=-4948>39-S9.6:*=(9.;*=1:2.T7*= = = &5745->9*= = C#= #>(47-.?*=CS8.(:1487':8(:1&.7*=

14 5 - ECTOMYCORHIZE

a= = Hyphes = externes = Réseau de Hartig = = Man- = teau = Hyphes externes = = 1 - ENDO 4 - ECTENDOMYCORHIZE Vési- = cule MYCORHIZE VESICULO- = ARBUSCULAIRE = Endoderme = Sporocyste Spororocyste = Peloton = Cylindre central lysé = 3 - ENDOMYCORHIZE A = PELOTONS (ERICACEES) = Pelotons = 2 - ENDOMYCORHIZE A PELOTONS (ORCHIDEES) = = = b = 1=47,&3.6:*=97*-&148*`=2&33.941`= ;.9&2.3*8 `=-47243*8_ = = -&25.,343 = 1&39*=, = = 1&39*=+ = = !&:`=*18=2.3S7&:=`= &?49* `= 5-485-47* `= ;.9&2.3*8`=-47243*8 = = B.,:7*= + _= 97:(9:7*= )*8= ).++S7*39*8= 2>(47-.?*8= &= *9= +43(9.433*2*39= )*8= S(-&3,*8= *397*= 1&= 51&39*= *9= 1*= (-&25.,343= )&38= 1*= (&)7*= )*= (*99*= 8>2'.48*= '_= '*8= 47,&3.82*8= &:949745-*8= 8439= +.,:7S8= *3= ;*79= *9= 1*8= -S9S749745-*8=*3=/&:3*`=)&57T8=*1488*`=,***_= = = 3)42>(47-.?*8= (942>(47-.?*8 = = *9= ;S8.(:148&7':8(:1&.7*= )%7(-.)S*8== )!7.(&(S*8== *(9*3)42>(47-.?*8=)!7.(&(S*8= = C#`= += ,= -= .`=/ = 142*742>(T9*8= &8.).42>(T9*8= 8(42>(T9*8= &8.).42>(T9*8=4:=8(42>(T9*8= = -&25.,3438=*9= 35,**=85_= 7-.?4(943.&=4:= *149.&1*8= 351***=85_= 342'7*=)*85T(*== 8(42>(T9*8= 7>48=*9=9*7.)45->9*8= %7(-.)S*8== !7.(&(S*8= 7'7*8=*9=!7.(&(S*8=&7':94)*8= 74:5*=)*= 3,.485*72*8= 5,/=***=85_= 5+***=85_= 5+****=85_= 51&39*8=*9= -*7'&(S*8=*9=&7'7*8= 342'7*=)*85T(*= 9745.(&:==5,,*=***=85_= 5S(.+.(.9S= B&.'1*= B&.'1*=A=+479*= B&.'1*=A=+479*= B&.'1*=A=+479*= &794:9=8&:+=7S,.438= &794:9=8&:+= .*25S7S`= 542.3&39=*3=7S,.438=9*25S7S*8= S5&79.9.43= 541&.7*8`=)42.3&39=*3= 7S,.438=541&.7*8= 94:3)7&= *9=8.*=):=:)8!89= ,S4,7&5-.6:*= 2.1.*:=9745.(&1= &8=)*=)S+472&9.43=)*=1&=7&(.3*= 5S+472&9.43=)*=1&=7&(.3*= #&39*&:=)*=2>(S1.:2=&:94:7= 97:(9:7*=)*= 7':8(:1*= *14943=.397&(*11:1&.7*= )*=1&=7&(.3*=*9=7S8*&:=)*= &79.,= 1&884(.&9.43= .397&(*11:1&.7*= = *397*=1*8=(*11:1*8=7&(.3&.7*8=(= = 5*149438=54:7=!7.(&(S*8= &'1*&:=+_= 5.;*78.9S=)*8=2>(47-.?*8=1*8=3:2S748=+439=7S+S7*3(*=A=1&=B.,_=+`=)&57T8=2.9-=]=$*&)`=,**2_=

15 _='&=2>(4-S9S749745-.*=*9=843=+43(9.433*2*39= = +_+ '&=)S(4:;*79*=):3*=34:;*11*=897&9S,.*=)-S9S749745-.*= = 03*=)*=(*8=51&39*8=).9*8=8&5745->9*8=&=+&.9=14'/*9=)*=342'7*:8*8=7*(-*7(-*8=*9=)*= )S'&98= &:= (4:78= ):= 66 T2* = 8.T(1*=a= .1= 8&,.9= ):3*= *7.(&(S*= 3438(-14745->11.*33*`= 1*= #4349745*= 8:(*85.3`= 4349745&= ->545.9->8= B.,_= ,_= !3= *++*9`= 03,*7= +2.*= &;&3O&.9= 6:*= (*11*8(.= S9&.9= *3= 7S&1.9S= 5&7&8.9*= )*8= 7&(.3*8= )*8= 5.38= A= (9S= )*86:*18= *11*= 54:88&.9`= 2&.8= 1&'8*3(*=)*=8:O4.7=*25U(-&.9=)*=;&1.)*7=(*99*=->549-T8*_=!3=+22,`=A&2.*380.=24397*=6:*= )*8= (-&25.,3438= 8*2'1*39= +472*7= :3= +*:97&,*= )*= 2>(S1.:2= &:94:7= )*8= 7&(.3*8= ):= #4349745*`= 8&38= 6:*= 1*8= 7&(.3*8= 3*= 5&7&.88*39= 5&7&8.9S*8= *9= (*8= 2U2*8= (-&25.,3438= 8*2'1*39=)S+472*7=)*8=7&(.3*8=574(-*8=)*= &&,:8=8>1;&9.(& _=1=)S24397*=5&7=(*8=4'8*7;&9.438= 6:*= 1*= #4349745*= 3*89= 5&8= 5&7&8.9*`= 2&.8= *89= 51:8= 574'&'1*2*39= &884(.S*= A= )*8= (-&25.,3438=6:.=1&=34:77.88*39_=1*=3*89=6:*3=+30*=6:*=/702&3=;&1.)*=(*99*=->549-T8*`= *3=:9.1.8&39=:3=2&76:&,*=&:= +. 1% ,=a=:3*=+4.8=&88.2.1S=5&7=1*=5.3`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`=*9=439=S9S=7S(*22*39=7*3422S*8=2>(4-S9S749745-*8= # =b='*&0*`=+33._==

B1*:7 =

,=(2 = !(&.11*8=343= (-14745->11.*33*8 =

41 = $&(.3*8=(4:79*8=*9= 9:'S7.8S*8`= *3)42>(47-.?S*8= >5-*8=*=9*73*8=)*=1*3)42>(47-.?* =

!(942>(47-.?* = =4398=2>(47-.?.*38==+472S8= )->5-*8=7*1.&39=1*8=)*:==2>(47-.?*8=

>5-*8=*=9*73*8=)*=1*(942>(47-.?* =

$&(.3*=):3=.3=;4.8.3 =

B.,:7*=, _= 4349745&=->545.9->8`= :3*=*7.(&(S*=2>(4-S9S749745-*=)&57T8=*1488*`=,***_=

16 = = 3)42>(47-.?*8= >(47-.?*8=+&(:19&9.;*8= (942>(47-.?*8= .58&(&1*8 89*7&1*8 3)42>(47-.?*8=) 7(-.)&(*&*= 5.&1*8 3)42>(47-.?*8=)7.(&(*&*= 6:.+41.&1*8 ======1439.*39=)*8=*85T(*8=# = 41&3&1*8 &2.&1*8 B&2.11*8= = *39.&3&1*8 *39.&3&(*&* = &884(.S*8=A=)*8= &77>&1*8 7.(&1*8 (-&25.,3438=a= 4349745&(*&* = 473&1*8 3)42>(47-.?.*38= &5.3)&1*8 (942>(47-.?.*38 = &1;&1*8 ;7&88.(&1*8 >79&1*8 B&,&1*8 :(:7'.9&1*8 48&1*8 B&'&1*8 =&1.)&1*8 &15.,-.&1*8 41>,&1&(*&*= *7&3.&1*8 = &=.+7&,&1*8 &39&1&1*8 &7>45->11&1*8 749*&1*8 &3:3(:1&1*8 *48.7.)&(*&*= E.3,.'*7&1*8 478.&(*&*= 422*1.3&1*8 4&1*8 7*(&1*8 &3)&3&1*8 &(&3)43.&(*&*= .1.&1*8 ;:72&33.&(*&*= .48(47*&1*8 85&7&,&1*8 = 1.82&9&1*8 7(-.)&(*&* = (47&1*8 *7&945->11&1*8 *9748&;.&(*&*= :2&,341..)*&* 7.:7.)&(*&*=

= B.,:7*= -_ = 11&88.+.(&9.43= 5->14,S3S9.6:*= )*8= 3,.485*72*8`= 57S8*39&39=54:7=(-&6:*=(1&)*=1*=9>5*= )42.3&39=)&884(.&9.43=2>(47-.?.*33*=&.38.=6:*=1*=9>5*=)42.3&39=)*8=57.3(.5&:==9&=438=(439*3&39= )*8= 2>(4-S9S749745-*8_= '*8= +&2.11*8= )3,.485*72*8= 2>(4-S9S749745-*8= 8439= 57S(.8S*8= 8:7= 1&= )74.9*=)*=1&7'7*`=*9=1*:7=(4:1*:7=.3).6:*=1*:7=9>5*=)*=2>(47-.?*$=8.=a=,74:5*=)3,.485*72*8= '&8&1*8= +472S= 5&7= 1*8= 2'47*11&1*8`= 1*8= 8>25-S&1*8`= 1*8= 11.(.&1*8`= 1*8= .7.2*3.&(S*8= *9= 1*8= :894'&11&>&1*8_=5&57T8=7:3)7*99`=,**,_ ==

17 +_, 5.;*78.9S=9&=4342.6:*=*9=7S5&79.9.43=,S4,7&5-.6:*=)*8= 51&39*8=2>(4-S9S749745-*8= = 1=*=.89*=51:8=)*=.**=*85T(*8=2>(4-S9S749745-*8=7S5&79.*8=*88*39.*11*2*39=)&38=-:.9= +&2.11*8= .3)S5*3)&39*8= *9= 21= ,*37*8= )&3,.485*72*8`= &:=6:*11*8= 8&/4:9*= :3= (1&)*= )-S5&9.6:*8=B.,_=-=b='*&0*`=+33._=1*8=*85T(*8=8439=57.3(.5&1*2*39=9745.(&1*8=.&'1*&:=,= 2&.8=)&:97*8=8439=7S5&79.*8=&:88.=*3=?43*=9*25S7S*`=(422*=1*8=*7.(&(S*8=*9=)*8=47(-.)S*8= .&'1*&:=,_='&=2&/47.9S=)*=(*8=*85T(*8=8439=47.,.3&.7*8=)8.*=):=:)8!89=*9=)2S7.6:*= ):=:)='*&0*`=+33.`=(477*8543)&39=A=)*8=54.398=(-&:)8=)*=).;*78.9S='.414,.6:*=#>*78= *9= &1_` =,***_=*:1*8=6:*16:*8=,*39.&3&(S*8`=97.:7.)&(S*8=*9=':72&33.(&(S*8=8439=57S8*39*8=*3= +7.6:*_='*8=*7.(&(S*8=8439=1*8=8*:1*8=A=U97*=).;*78.+.S*8=*3=7S,.43=9*25S7S*`=41=8*=974:;*39= &:88.=:3*=6:.3?&.3*=)*85T(*8=)47(-.)S*8=2>(4-S9S749745-*8_== = B&2.11*= 842'7*=)*85T(*8= &'.9&9=4:=?43*= )&3,.485*72*8= 2>(4-S9S749745-*8= ,S4,7&5-.6:*= %7(-.)S*8= +1+= .*25S7S=*9=.745.(&1= :72&33.&(S*8= 3*= .745.(&1= .7.:7.)&(S*8= 1-= .745.(&1= *39.&3&(S*8= .*= 2S7.6:*= ):= :)= *9= +7.6:*= 1478.&(S*8= ,1= 2S7.6:*= ):= :)= *9= 8.*= != !7.(&(S*8= +/= .*25S7S= 41>,&1&(S*8= ,= 3)43S8.*= *9748&;.&(S*8= ,= 8.*=):=:)8*89= '&(&3)43.&(S*8= += #*=.6:*= *48.7.)&(S*8= += #&)&,&8(&7=

= &'1*&:=,_= $S5&79.9.43=8>89S2&9.6:*=*9=,S4,7&5-.6:*=)*8=*85T(*8=2>(4-S9S749745-*8=5&72.=1*8= &3,.485*72*8_=5&57T8='*&0*`=+33._ = = '*= 24)*= )*= ;.*= 2>(4-S9S749745-*= (-*?= (*8= +&2.11*8= 8*2'1*= U97*= &55&7:= )*= +&O43= .3)S5*3)&39*`=)&38=94:8=1*8=S(48>89T2*8`=9*25S7S=4:=9745.(&:=_=1*79&.3*8=+&2.11*8=(422*= 1*8= ':72&33.(&(S*8= 8439= (4251T9*2*39= 2>(4-S9S749745-*8`= 1&.88&39= 8:5548*7= :3= 8*:1= S;S3*2*39`= &3(.*3`= )&55&7.9.43= )*= 1&= 2>(4-S9S749745-.*_= 1-*?= 1*8= 47(-.)S*8= 1&= 2>(4-S9S749745-.*= 8*7&.9= &55&7:*= .3)S5*3)&22*39= ,*= +4.8= B.,_= 1`= #41;7&>`= ,***`= *9= 2U2*=51:8.*:78=+4.8=&:=8*.3=):3*=97.':=(422*=(*11*=)*8=3*499.S*8=B.,_=3_=1=3*=.89*=5&8= )*= 51&39*= 2>(4-S9S749745-*= +488.1.8S*= (433:*`= (*= 6:.= *25U(-*= 1&= )&9&9.43= )*= (*8= S;T3*2*398_=1*5*3)&39`=1S9:)*=)*8=':72&33.(&(S*8`=7S5&79.*8=A=1&=+4.8=84:8=1*8=8S48=*9= &1S489745.6:*8`= 8:,,T7*= :3*= 8*:1*= &55&7.9.43= &39S7.*:7*= A= 1&= 8S5&7&9.43= *397*= (*8= (439.3*398`= 8.9:S*= ;*78= 8+-*= 2.11.438= )&33S*8_= 1*99*= +&2.11*= 8*7&.9= 47.,.3&.7*= )*8= 8S49745.6:*8`= *9= &:7&.9= 2.,7S= ;*78= 1*89= &;&39= )*= 8*= ).;*78.+.*7= 84:8= 1*8= &1S49745.6:*8`= 57.3(.5&1*2*39=A=1%1.,4(T3*`=:3*=5S7.4)*=6:.=&=;:=1*=9*38.43=):=(4:;*79=+47*89.*7=)&38= (*8=7S,.438=#*70=8= *9=&1_` =,**2_=

18 55&7*.1=7*574):(9*:7 = B7:.98=574):.8&39=)*=342'7*:8*8= 5*9.9*8=,7&.3*8= 7&.3*8=)S54:7;:*8=)*=7S8*7;*8`= ,*72.3&9.43=)S5*3)&39*=):= (-&25.,343=3:97.9.43= = /*= 2 = B1*:78=A=+147&.843=(4:79*`=A=9*3)&3(*= &:94,&2*=4:=5411.3.8&9.43=5&7=)*8= .38*(9*8=,S3S7&1.89*8= = B1*:78=)*=1&=2U2*=(4:1*:7=6:*=1&=9.,*=

55&7*.1=(&:1. 3&.7* = *79*=)*8=(-14745->11*8= 14:1*:7=842'7*= B*:.11*8=7S):.9*8=A=)*8=S(&.11*8= $S):(9.43=):=342'7*=)*=8942&9*8`=4:= &'8*3(*= -14T2*=7S):.9`=5*:=)*=8(1S7*3(->2*=

55&7*.1=7&(.3&.7* = $&(.3*8=(4:79*8=4:=97T8=7&2.+.S*8= $-.?42*=->5*79745-.S= .7T8=+479=9&:==)*=2>(47-.?&9.43= 94(0&,*=)&2.)43= 1&5&(.9S=)*=)472&3(*=4:=574):(9.43= )*=9.,*8=84:9*77&.3*=.25479&39*= $*574):(9.43=;S,S9&9.;*=5&7=1*= 7-.?42*=

+=(2 =

= = = = = B.,:7*=._= '*8=8>3)72*8=)*8=51&39*8=2>(4-S9S749745-*8`=.11:897S8=(-*?=:3*=47(-.)S*`= $*499.&=3.):88&;.8 =2&,*=)*=1&=B147*=)*=1&71==*1=#&,3:8='.3)2&3`=+3*/_==

19 '&55&7.9.43=)*8=47(-.)S*8`=:3*=+&2.11*=*897U2*2*39=).;*78.+S*=&;*(=57T8=)*=-*=***= *85T(*8= 17.''= *9= &1 _`= ,**-= &= S9S= 7S(*22*39= 7*54:88S*= )*= 8-*= A= 82*= 2.11.438= )&33S*8= $&2.7*?= *9=&1_` =,**1`=;.*.11.88&39=)*=6:*16:*8=/*=2.11.438=)&33S*8=(*99*=+&2.11*=/:86:*=1A= (438.)S7S*= (422*= 7S(*39*_= := 8*.3= )*= (*99*= +&2.11*`= 1&= ).897.':9.43= )*8= S;S3*2*398= )&55&7.9.43=)*=1&=2>(4-S9S749745-.*=(43):.8*39=A=)*8=-.894.7*8=S;41:9.;*8=97T8=).++S7*39*8= *9= 51:8= 4:= 24.38= &3(.*33*8= 8*143= 1*8= ,74:5*8`= (422*= *3= 9S24.,3*= 1S9:)*= )*= 6:*16:*8= ,T3*8= (-147451&89.6:*8= a= &1478= 6:*= 1&= 51:5&79= )*8= ,T3*8= (-147451&89.6:*8= 3*= 8439= 5&8= &251.+.&'1*8=(-*?=1*8=,&8974).S*8`=.18=1*=8439=(-*?=1&=51:5&79=)*8= 47&1147-.?& =855_=&77*99=]= B7*:)*389*.3`=,**2_= = = +_- '*8=97&.98=)S7.;S8=)*8=2>(4-S9S749745-*8= = .*3= 6:*11*8= &55&79*3.*33*39= A= )*8= +&2.11*8= 97T8= ).++S7*39*8`= (*79&.38= 97&.98= '.414,.6:*8=8439=(422:38=A=94:9*8=1*8=51&39*8=2>(4-S9S749745-*8_=.4:9=)&'47)`=(*=24)*= )*=;.*=-S9S749745-*=1*:7=&=5*72.8=)4((:5*7=)*8=-&'.9&98= &=57.47. =)S+&;47&'1*8=&:==51&39*8= 5-4948>39-S9.6:*8`=(422*=84:8=1*=(4:;*79=+47*89.*7=9745.(&1=4:=9*25S7S='*&0*`=+33._=1*8= 84:88'4.8= 8439= 84:;*39= 2&76:S8= 5&7= 1&= 57S8*3(*= 8.2:19&3S*= )*= 51:8.*:78= *85T(*8= )*= 51&39*8= 2>(4-S9S749745-*8`= :3= 9&:== )-:2.).9S= &88*?= .25479&39= *9= :3= (4:;*79= -*7'&(S= 84:;*39=1.2.9S_=5&:97*=5&79`=1*8=*85T(*8=2>(4-S9S749745-*8=8*=).89.3,:*39=+&(.1*2*39=)*8= &:949745-*8=5&7=1*:7=2475-414,.*=a=1*:78=+*:.11*8=8439=84:;*39=7S):.9*8=A=)*8=S(&.11*8=*9=3*= 5488T)*39= 5&7+4.8= 3.= 8942&9*8= 3.= (-14745->11*`= 9&3).8= 6:*= 1&= 51&39*= &7'47*= 84:;*39= :3*= 5.,2*39&9.43=+43(S*=B.,_=,`=.=b='*&0*`=+33.=b=$&82:88*3`=+33/_='&=9.,*=5488T)*=:3=8>89T2*= ;&8(:1&.7*= 24).+.S`= 41= 1*= 5-14T2*= *89= 7S):.9= &.38.= 6:*= 1*8= (47)438= )*= 8(1S7*3(->2*_= !3= ,S3S7&1`= (*8= 51&39*8= 439= :3= &55&7*.1= (&:1.3&.7*= S,&1*2*39= 7S):.9`= 6:.= 3&55&7&9= 6:&:= 242*39= )*= 1&= +147&.843_= '*= 342'7*= )*= +1*:78= 5&7= .3).;.):= *89= &88*?= ;&7.&'1*= 8*143= 1*8= *85T(*8_= '&= 5411.3.8&9.43`= A= 9*3)&3(*= &:94,&2*`= 5*:9= &:88.= U97*= &88:7S*= 5&7= )*8= .38*(9*8= ,S3S7&1.89*8= '*&0*`= +33.=b= #41;7&>`= ,***_= '*8= +7:.98= 574):.8*39= :3= 342'7*= )*= ,7&.3*8= '*&:(4:5= 51:8=S1*;S= 6:*=(-*?= 1*8=9&=438= &:949745-*8= 1*8= 51:8= 574(-*8_=1*8= ,7&.3*8= 8439= )S54:7;:*8=)*=7S8*7;*8=&:=24.38=(-*?=1*8=,*39.&3&(S*8`=1*8=*7.(&(S*8=*9=1*8=47(-.)S*8=*9= 8439=)*=9&.11*=*=97U2*2*39=7S):.9*=)*=147)7*=)*=/* 2`=(*=6:.=*=51.6:*=1*:7=6:&1.+.(&9.+= )*==,7&.3*8854:88.T7*8=== ):89=8**) ==b=7).99.`=,***=b=B.,_=/_= .= 1*8= 5&79.*8= &S7.*33*8= )*= (*8= 51&39*8= 8439= 7&7*2*39= ;.8.'1*8= *3= 84:8= '4.8`= 1*8= 5&79.*8= 84:9*77&.3*8= 8439= *3= 7*;&3(-*= *=97U2*2*39= )S;*1455S*8= *9= 24).+.S*8`= 6:.1= 8&,.88*=)*=7&(.3*8=4:=)*=7-.?42*8='*&0*`=+33.=b=$&82:88*3`=+33/=b=2-4+`=,**-_=1=&77.;*= 2U2*= 6:*= 1&= 9.,*= +1*:7.88*= 84:8= 9*77*= (422*= (-*?= 5.54,.:2= &5->11:2 `= 6:.= 5&7= &.11*:78= 5*:9=7*89*7=*3=)472&3(*=84:8=9*77*=5*3)&39=57T8=)*=-*=&38=:22*7-&>*8`=+3/+=b=4>7.30.`= +321=b=$4'.3`=+323_=1*99*=57S)42.3&3(*=)*=1&=;.*=84:9*77&.3*=*9=1->5*79745-.*=)*8=7&(.3*8= 3*=(477*8543)=5&8=9&39=A=1&=2.8*=*3=51&(*=)47,&3*8=)*=7S8*7;*8`=2U2*=8.=:3=894(0&,*=>=*89= 7S&1.8S= 84:8= +472*= )&2.)43_= 1= 8&,.9= )&;&39&,*= ):= )S;*1455*2*39= )47,&3*8`= 349&22*39=1*=7-.?42*`=)439=1&=51:5&79=)*8=(*11:1*8=8439=(4143.8S*8=5&7=1*=(-&25.,343=)*= +&O43=):7&'1*=*9=+472*39=)*8=2>(47-.?*8$=B.,_=+_='*8=2>(47-.?*8`=4'8*7;S*8=(-*?=2/=)*8= 51&39*8= 9*77*897*8`= .251.6:*39= )*8= (-&25.,3438= 97T8= ).;*78= .&'1*&:= +`= 7*(*;&39= ):= (&7'43*= 47,&3.6:*= )*= 1&= 51&39*= -9*_= 1*5*3)&39`= )&38= 1*= (&8= )*8= 51&39*8= 2>(4-S9S749745-*8`= (*11*88(.= 3*= 5*:;*39= +4:73.7= )*= 8:(7*8= .88:8= )*= 1&= 5-4948>39-T8*`= (*= 6:.=84:1T;*=)*=342'7*:8*8=6:*89.438=6:&39=&:=+43(9.433*2*39=)*=(*99*=8>2'.48*_==

20 3(&)7S=+_= S9-4)*8=(74.8S*8=).)*39.+.(&9.43=)*8=(-&25.,3438=2>(47-.?.*38=*9=)*=1&== 84:7(*=)*=(&7'43*=47,&3.6:*=(-*?=1*8=*85T(*8=2>(4-S9S749745-*8_= 251.+.(&9.43=5&7= = *8:7*=):=) +- =*9=) +/ = '*8=*85&(*:7=.39*7,S3.6:*8=.39*73*8=)*=158= '*= (&7'43*= *9= 1&?49*= 8439= 57S8*398= )&38= 7.'4842&1= 3:(1S&.7*= 8439= )*8= 7S,.438= 343= 1 &92485-T7*= 84:8= +472*= )*= ).++S7*398= (4)&39*8= 57S8*39*8= *3= /*= (45.*8= )&38= 1*= .84945*8= +- 1`= +, 1`= +/ 8= *9= +. 8_= '*8= *3?>2*8= ,S342*_= '&= ).,*89.43= )*8= 574):.98= 1$= 5&7= 8439= (&5&'1*8= )*= ).8(7.2.3*7= 1*8= .84945*8= )*8= *3?>2*8= $B'= 4:= 1*:7= 8S6:*3O&,*= 5&7= 1*:7= 54.)8= 241S(:1&.7*= *9= 1*8= .84945*8= 7S;T1*= 1*= 541>2475-.82*= )*= 1._= 1*= 51:8= 14:7)8= 8439= 24.38= &88.2.1S8_= '*8= +7&,2*39= 57S8*39*= :3*= +&.'1*= ;&7.&9.43= &:949745-*8= 439= :3*= 8.,3&9:7*= .84945.6:*= .397&85S(.+.6:*=0-=*9=5*72*9=)*=).89.3,:*7= ).++S7*39*= )*8= (-&25.,3438= !1#`= 6:.= 8*= 1*8= *85T(*8_= 5*8= &247(*8= 85S(.+.6:*8= )*8= ).89.3,:*39= &:88.= )*8= 8&5745->9*8`= 6:*= (-&25.,3438= (422*= .+B= *9= ..= 143=(425&7*=A=1&.)*=):= +- 1=*9= +/ 8=5&7= 5*72*99*39= )*= 3&251.+.*7= 6:*= 158= ):= 7&55479= A= :3*= 7S+S7*3(*= .39*73&9.43&1*`= :3= (-&25.,343= 2>(47-.?.*3_= 5*8= '&36:*8= )*= (&7'43&9*= 4(S&3.6:*= 54:7= 1*= 1= *9= 1*= 8 ,= )433S*8= 5*72*99*39= )*= (425&7*7= 1.= ):= &92485-S7.6:*=54:7=1*=8= (-&25.,343= *9= )*= 1.)*39.+.*7= &7)*8= ]= = 7:38`=+33-_= +- 1- +- 12 +, 147(-.)S*8 +- 12 +, 17S+S7*3(* = +B = . = ==== +- 12 +, 17S+S7*3(*= = +2 = + = /_2 = ,/ = '*8= -S9S749745-*8= 439= 1&= 8.,3&9:7*= .84945.6:*= )*= 1&= 2&9.T7*= 47,&3.6:*= 6:.18= 439= &'847'S= 54:7= 1*= 1`= *9= :3*= 8.,3&9:7*= S6:*3(*=(4)&39*`=).89.3,:*=1*8=47)7*8= *37.(-.*=54:7=1*=85&<843= *9=&1_` =,**,_=!3= +- +/ S6:*3(*=343=(4)&39*`=).89.3,:*=1*8=*85T(*8= 2*8:7&39= 1*= 1= *9= 8=)*8=*85T(*8=# `= *9= *3= 1*= (425&7&39= A= )&:97*8= 47,&3.82*8= ):= 2U2*= *3;.7433*2*39`= 43= 5*:9= &.38.= 8&;4.7= )*= 6:4.= (*8= *85T(*8= 8*= 34:77.88*39= 1425&7&.843=&:=='&36:*8=)*= (+_ &79.(1*= _== )433S*8=4:=&:==.=)*(942>(47-.?*8= +/ 8=*3== +, %7(-.)S*=# = )&7'7*8=574(-*8= +* !1# = 2 0 !1# = . -&25.,343= = , * :949745-*= 8, 8. 8-. 8-, 8-* 8,2 8,0 8,. 8,, :949745-* = +- 1=*3= =

!1# = %7(-.)S*=

!1# =

B1:==)*=1=47,&3.6:*= B1:==)*&:== *9=8*18=2.3S7&:==8`==

= =

21 = +_. '&=3:97.9.43=)*8=51&39*8=2>(4-S9S749745-*8= = ;&39= 2U2*= )*= (433&97*= 1.)*39.9S= *=&(9*= )*8= (-&25.,3438= 2>(47-.?.*38`= 1*= 24)*= )*= 3:97.9.43= )*= (*8= 51&39*8= &;&.9= S9S= S1:(.)S_= '*=5S7.*3(*= 8S2.3&1*= )*= /702&3= +30*= &= *3= *++*9= )S24397S= 1*= 97&38+*79= )*= (&7'43*= 7&).4(&9.+= )*= .3:8= 2:7.(&9& `= ;.&= 1*= 2>(S1.:2=):3=(-&25.,343`=A= 4349745&=->545.9->8 _=1:8=9&7)`=#(A*3)7.(0= *9=&1_` =,***= 439=24397S=1*=97&38+*79=)*=(&7'43*=7&).4&(9.+=)*= #&1.==7*5*38 =&:==51&39:1*8=)*= 47&1147-.?&= 2&(:1&9& _='*=8:.;.=):=(&7'43*=7&).4&(9.+=3*89=5&8=1*=8*:1=.3).(&9*:7=)*=1&=84:7(*=)*=(&7'43*= 47,&3.6:*= :9.1.8S= 5&7= (*8= 51&39*8_= !3= *++*9= 1&= 2*8:7*= ):= +- 1= *9= ):= +/ 8= =1&= 8.,3&9:7*= .84945.6:*== &= 5*72.8= )S9*3)7*= 1*8= 7*(-*7(-*8= .3= 8.9: `= *3= )*-478= )*8= 2.(74(482*8= )*= 1&'47&94.7*= !3(&)7S= +_= 4349745&= ->545.9->8 = *9= $*499.&= 3.):88&;.8 `= (422*= 94:9*8= 1*8= 51&39*8= 2>(4-S9S749745-*8= 9*25S7S*8= 439= :3= +- 1= .)*39.6:*= A= (*1:.= )*= (-&25.,3438= *(942>(47-.?.*38=!1#`=B.,_=+_=1422*=(*8=(-&25.,3438=3*=8439=5&8=&:949745-*8=*9=8439= .251.6:S8=)&38=:3*=8>2'.48*=*(942>(47-.?.*33*=&;*(=)*8=&7'7*8`=6:.=1*:7=+4:73.88*39=)*8= 8:(7*8`=(*8=&7'7*8=(4389.9:*39=)43(=54:7=1*8=)*:==*85T(*8=)*=51&39*8=2>(4-S9S749745-*8=1&= 84:7(*=)47.,.3*=):=(&7'43*=47,&3.6:*`=&88.2.1S* =;.&= 1*=(-&25.,343=!3(&)7S=+`= *'&:*7= ]=#*>*7`=,**-=b=.7:)*11= *9=&1_` =,**-_='&=+472*=)*=97&38+*79=):=(&7'43*=3*89=5&8=(433:*`=.1= 54:77&.9=8&,.7=)*=2&33.941=4:=)*=97*-&148*=2.9-=]=$*&)`=,**2=*9=1*=9*72*=)*=(&7'43*= 47,&3.6:*=8*7&=:9.1.8S=54:7=)S(7.7*=1&=2&9.T7*=47,&3.6:*=97&38+S7S*_== 5*= 2U2*= 1&= 24)&1.9S= ):= 97&38+*79= 3*89= 5&8= (433:*_= 1-*?= 1*8= 47(-.)S*8`= 1*= 2>(S1.:2=):=(-&25.,343=+472*=)*8=5*149438=)&38=1*8=(*11:1*8=)*8=7&(.3*8=8&38=+7&3(-.7=1&= 2*2'7&3*= 51&82.6:*= B.,_= +`= B.,_= /_= 1*8= 5*149438= 8439= 1>8S8= )*= +&O43= 7S,:1.T7*= )&38= (-&6:*=(*11:1*`=(*=6:.=(43):.9=A=:3=(4389&39=7*34:;*11*2*39=)*=1.39*7+&(*=2>(47-.?.*33*= *9= )*8= ;&7.&9.438= 85&9.&1*8= *9= 9*2547*11*8= )*= 1&= 2>(47-.?&9.43_= '*= 97&38+*79= )*= (&7'43*= 47,&3.6:*=5*:9=&;4.7=1.*:=5*3)&39=6:*=1*=2>(S1.:2=*89=;.;&39=*9=1478=)*=1&=1>8*=.7:)*11=*9= &1_`=,**-_=1-*?=1*8=*7.(&(S*8`=(*8=5*149438=3*=8439=5&8=1>8S8`=*9=1*=97&38+*79=8*=+&.9=)43(=):= ;.;&39=):=(-&25.,343=.*)*7844= *9=&1 _`=,**1&_= &7= &.11*:78`= 1*8= *85T(*8= (-14745->11.*33*8= )47(-.)S*8= *9= )*7.(&(S*8= 57S8*39*39= &:88.= )*= 9*18= 5*149438_= 1-*?= 1*8= *85T(*8= (-14745->11.*33*8`= 1*= 8*38= )*8= +1:== )*= (&7'43*= 47,&3.6:*= 5*:9= ;&7.*7_= = 1S9&9= )*= ,7&.3*`= 94:9*8= 1*8= 47(-.)S*8= 7*O4.;*39= ):= (&7'43*= )*= 1*:78=(-&25.,3438=2>(47-.?.*38`=*11*8=8439=)43(=94:9*8=2>(4-S9S749745-*8_=!3=7*;&3(-*`= A=1B,*=&):19*`=*11*8=+4:73.88*39=):=(&7'43*=47,&3.6:*`=.88:=)*=1*:7=5-4948>39-T8*`=A=1*:78= (-&25.,3438= 2>(47-.?.*38_= .4:9*+4.8`= 1*:78= (-&25.,3438= 2>(47-.?.*38= 5*:;*39= *3(47*= 1*:7= +4:73.7= ):= (&7'43*=47,&3.6:*=1&2*743= *9=&1 _`= ,**0`= 2&.8= (*9= &55479= 3*=7*57S8*39*= 6:*= += A= -= = )*= 1*:7= ':),*9= (&7'43S= 949&1_= 1*79&.3*8= 47(-.)S*8= *9= *7.(&(S*8= (-14745->11.*33*8=7*O4.;*39=:3*=6:&39.9S=)*=(&7'43*='.*3=8:5S7.*:7*=/:86:A=2/=)*=1*:7= ':),*9=(&7'43S`= *'&:*7=]=#*>*7`=,**-=*9=3*=+4:73.88*39=5&8=)*=(&7'43*=&:=(-&25.,343_= 1*8= )*73.T7*8= :9.1.8*39= )43(= )*:== 84:7(*8= )*= (&7'43*=a= ):= (&7'43*= .347,&3.6:* = ;.&= 1&= 5-4948>39-T8*=*9=):=(&7'43*=47,&3.6:*= ;.& =1*:78=(-&25.,3438`=*9=8439=).9*8=2.=49745-*8$_= = 1*99*= &5574(-*= &= 5*72.8= )S9:).*7= 1*= 89&9:9= 9745-.6:*= )*= 342'7*:8*8= 51&39*8= 2>(4-S9S749745-*8`=&.38.=6:*=)*=1*:78=574(-*8=5&7*398=(-14745->11.*38`=2&.8=1*=(429=)*=(*8= 2*8:7*8=&=1.2.9S=1*=342'7*=)S9:)*8_== = =

22 -&25.,343=2>(47-.?.*3= (414,.*= S9-4)*==E43*=)S9:)*= S+S7*3(*== 7(-.)S*8=2>(4-S9S749745-*8=)*=2.1.*:=9*25S7S= *5-&1&39-*7&= .&>147=]=7:38`= .-S1S5-47&(S*8= &:89.3&*= +331= 47&1147-.?&= .&>147=]=7:38`= 2&(:1&9&=*9= $:88:1&(S*8= 0= +333= _=2*79*38.&3&= .=A=5&79.7=)*8= = = **=&1*(97.8= 7&(.3*8=4:=):= .&>147= *9=&1_` =,**-= 85.(&9&= *'&(.3&1*8= 7-.?42*= !1#= = *_=7*;41:9&= = = *1488*= *9=&1_` = ,**,'`= $*499.&=3.):88&;.8= *'&(.3&1*8= B7&3(*= #(A*3)7.(0= *9=&1_` = ,***= Z*12*7=]=1:77&-`= 47&1147-.?&= 3(433:=8:7=1*=8.9*= :.;.=)*=(&7'43*= 0= +33/=b=#(A*3)7.(0= 97.+.)&= 9-S1S5-47&(S*8= 7&).4&(9.+= *9=&1_` =,***= 7(-.)S*8=2>(4-S9S749745-*8=)*=2.1.*:=9745.(&1= -&2&*,&8974).&= &&,&2*= *9=&1_` = *7&94'&8.).:2 =85_= 8.*= 8.040.&3&= ,**2'= .=A=5&79.7=)*= ,.54).:2= 5*&73&1*>=]='*= $:88:1&(S*8= (:19:7*=)*8=5*149438= ;&7.*,&9:2= !1#= :897&1.*= 74(6:*`=,**0= = ,.54).:2= = = 5*&73&1*>=]='*= $:88:1&(S*8= -&2.1943.&3:2= 74(6:*`=,**0= --.?&39-*11&= "-&3&9*5-47:8= %'8*7;&9.43= :897&1.*= F&7(:5`=+32/= ,&7)3*7.= ,&7)3*7.======&2&)&`=+3-3`= .&1*41&= .=A=5&79.7=)*= .*7&8-.9&`=+32/`= 72.11&7.&=2*11*&= <&543= 8*59*397.43&1.8= (:19:7*=)*8=5*149438= .*7&8-.9&=*9= = 1-:2&3`=+321= .&1*41&=->)7&= &42*8 =85_== = %'8*7;&9.438=8:7=):= -.1.55.3*8= :7,*++`=+3/3= .&8974).&=*1&9&= 72.11&7.& =85_= '4.8=2479`=4:= <&543= A:8&34`=+3++= .&8974).&= ).897.':9.43= 84:;*11*= &42*8 =85_= 1&25'*11`=+30.= 8*8&24.)*8= ,S4,7&5-.6:*= ZS1&3)*= .&8974).&= = 0*749:8 =85_= %'8*7;&9.438= 3)43S8.*= :7,*++`=+3/3= /&;&3.(&= = .&8974).&=8.2.1.8= -*8.3.(.:2= 85_= = .=A=5&79.7=):= '&=$S:3.43= #&7948= *9=&1 _`=,**3= 7-.?42*======&&2&94= *9=&1_` = 5.54,.:2= .=A=5&79.7=)*= 1457.3&(S*8= <&543= ,**/=b=&&,&2*= *9= 748*:2= (:19:7*=)*8=5*149438= &1_`= ,**2&= :145-.&= .=A=5&79.7=)*8= <&543`= %,:7&8.8:/.9&=]= 1457.3&(S*8= ?411.3,*7.= 7&(.3*8= #>&32&7= &:0&<&`=,**2'= .*898=)*= :7,*++`=+3-,`= ,.)>2451*=.8= 85_= &7&82.:8 =85_= 3)43S8.*= ,*72.3&9.43= +3-0`=+3/3= :7.(:1&7.& =85_= = <&543`= 7>9747(-.8= 02&9&`=+331`= *39.3:1& =85_= 84:;*11*= 4(-4'.*38.8= +332&`=+332'= "7&2*9*8 =85_= = :.3S*= &2&)&=]= .&1*41&=&19.88.2&= 7>9-742>(*8 =85_= <&543= 8&0&2:7&`=+30-= 1:118(-&*,*1.&= >(*3&= 85_`= = :&)*14:5*= #&7948= *9=&1 _`=,**3= &5->11&= .>2345:8= 85_= 24&3.&=&:897&1.8= >(45*7)43 =85_= %'8*7;&9.438= 0= 1&25'*11`=+31*= = &'1*&:=-_= 5.;*78.9S=)*8=(-&25.,3438=2>(47-.?.*38=5&79*3&.7*8=)47(-.)S*8=2>(4-S9S749745-*8_=

23 +_/ )*39.9S=)*8=(-&25.,3438=&884(.S8=&:==2>(4-S9S749745-*8= = '&=(:19:7*=)*8=(-&25.,3438=)*8=2>(4-S9S749745-*8=8*89=143,9*258=841)S*=5&7=)*8= S(-*(8= 7&82:88*3`= +33/= *9= (*= 8439= 1*8= &;&3(S*8= *3= '.414,.*= 241S(:1&.7*`= 349&22*39= 1:9.1.8&9.43= )&247(*8= 85S(.+.6:*8= &:== (-&25.,3438`= 9*11*8= (*11*8= )*8= .= 39*73&1= .7&38(7.'*)= 5&(*7= )*= 158= 7.'4842&1`= !3(&)7S= +`= 6:.= 439= 5*72.8= ).)*39.+.*7= 1*8= (-&25.,3438=&884(.S8_='*8=(433&.88&3(*8=8:7=1*:7=S(414,.*=5*72*99*39=).3+S7*7=1&=5488.'1*= 84:7(*=):=(&7'43*=7*O:=5&7=1*8=51&39*8=2>(4-S9S749745-*8_== '&= 51:5&79= )*8= *85T(*8= 2>(4-S9S749745-*8= )*= ,*39.&3&(S*8`= 97.:7.)&(S*8`= ':72&33.&(S*8= *9= 9-.82.&(S*8= 8439= &884(.S*8= A= )*8= ,142S742>(T9*8= .)&7943)4= *9= &1_` = ,**,=b= '*&0*`= ,**.=b= B7&30*= *9= &1_` = ,**0`= (422*= 1*:78= 574(-*8= 5&7*398= &:949745-*8= 7:3)7*99`=,**,=b=.&'1*&:=+`=B.,_=+_=.4:9*8=(*8=*85T(*8=2>(4-S9S749745-*8=8439=&884(.S*8= 97T8= 85S(.+.6:*2*39= A= :3= (1&)*= )*= (-&25.,3438=a= (-&6:*= 51&39*= *89= &884(.S*= A= :3= 8*:1= (-&25.,343`=*9=94:8=1*8=51&39*8=)*=1&=2U2*=*85T(*=8439=&884(.S*8=A=)*8=(-&25.,3438=)*=1&= 2U2*= *85T(*= 4:= ):= 2U2*= (1&)*= .&>147= *9= &1_` = ,**,_= '*8= ,142S742>(T9*8= +472*39= )*8= *3)42>(47-.?*8= &;*(= 1*8= 7&(.3*8= )*= 51&39*8= -*7'&(S*8= *9= )*= 1&= 51:5&79= )*8= &7'7*8= 9745.(&:==b=.18=7*O4.;*39=):=(&7'43*=47,&3.6:*=&88.2.1S=5&7=1*:78=-9*8=B.,_=+=*9=.&'1*&:= +_='*8=51&39*8=2>(4-S9S749745-*8=&884(.S*8=A=(*8=2U2*8=,142S742>(T9*8=7*O4.;*39=)43(= ):=(&7'43*=.3.9.&1*2*39=574):.9=5&7=)*8=&:949745-*8`=(422*=57S(S)*22*39=)S(7.9_= '*8= *7.(&(S*8= 2>(4-S9S749745-*8= 8439= &884(.S*8= 97T8= 85S(.+.6:*2*39= A= )*8= (-&25.,3438= *(942>(47-.?.*38`= 9*18= 6:*= )*8= 7:88:1&(S*8= *9= )*8= 97.(-4142&9&(S*8= .)&7943)4`= ,**/`= 6:.= 1*:78= +4:73.88*39= ):= (&7'43*= 574):.9= 5&7= 1*8= &7'7*8= /702&3`= +30*_= '*8= *7.(&(S*8= &:949745-*8= 8439= *3= 7*;&3(-*= &884(.S*8= A= )*8= (-&25.,3438= &55&79*3&39= A= 147)7*= )*8= -*149.&1*8= )*8= &8(42>(T9*8`= 2&.8= 5&7+4.8= &:88.= )*8= '&8.).42>(T9*8=*(942>(47-.?.*38=54:7=1*8=(1&)*8='&8&:==)*=(*99*=+&2.11*=.&'1*&:=+_= .4:9*8=1*8=47(-.)S*8=57S8*39*39=:3=2U2*=9>5*=)&884(.&9.43=2475-414,.6:*2*39= 8.2.1&.7*= A= (*1:.= )*8= *7.(&(S*8_= '*8= *85T(*8= 2>(4-S9S749745-*8= 8439= &884(.S*8= 97T8= 85S(.+.6:*2*39= A= :3*= 8*:1*=*85T(*= )*=(-&25.,343=*(942>(47-.?.*3_= !3=2.1.*:= 9*25S7S`= 974.8= +&2.11*8= )42.3*39= *3= 9*72*8= )*= +7S6:*3(*= )&884(.&9.43=a= 1*8= 8*'&(.3&1*8`= 1*8= 9-S1S5-47&(S*8= *9= 1*8= 7:88:1&(S*8= .&'1*&:= -=b= .&>147= *9= &1_` = ,**,=b= 5*&73&1*>`= ,**1_= 1*5*3)&39`= 1&= 51:5&79= )*8= 47(-.)S*8= 2>(4-S9S749745-*8= 8439= 9745.(&1*8`= *9= (*11*88(.= 439= 7S;S1S= /:86:A= 57S8*39`= :3*= &884(.&9.43= 94:/4:78= 97T8= 85S(.+.6:*= 2&.8= &;*(= )*8= (-&25.,3438= 8&5745->9*8= 4:= 5&7&8.9*8= .&'1*&:= -_= '*8= 47(-.)S*8= &:949745-*8= 8439= &884(.S*8= A= )*8= (-&25.,3438= (4251T9*2*39= ).++S7*398`= :3= *38*2'1*= 541>5->1S9.6:*= 7&88*2'1S=84:8=1*=9*72*=)*=7-.?4(943.&$`=&:88.=(433:8=(422*=8&5745->9*8=4:=5&7&8.9*8_=18= +472*39=:3*=8>2'.48*=*9=)439=1*8=47(-.)S*8=)S5*3)*39=)T8=1*:7=,*72.3&9.43=!3(&)7S=,=b= .&'1*&:=.=b=$&82:88*3=+33/_='&=85S(.+.(.9S=)*=(*99*=8>2'.48*=+&.9=94:/4:78=)S'&9=a= .3=;.974 `= 1&= ,*72.3&9.43= *89= 84:;*39= 5488.'1*= &;*(= :3*= 2:19.9:)*= )*= (-&25.,3438= 1:79.8`= +3-1=b= +3-3=b= &)1*>`=+31*`=#&8:-&7&=]=A&98:>&`=+323=b=+33+=b=#&8:-&7&= *9=&1_` =+33-=9&3).8=6:*= .3= 8.9: `= 1*= 342'7*= )*= 5&79*3&.7*8= 2>(47-.?.*38= *89= 51:8= 7S):.9= 9&39= (-*?= 1*8= ,7&.3*8= #(= 1472.(0= *9= &1_` = ,**.= 6:A= 1B,*= &):19*= %9*74= *9= &1_` = ,**._= 5*= 51:8= 1*= +&.'1*= 8:((T8= )*= ,*72.3&9.43=84:1.,3*=(*99*=85S(.+.(.9S=&:=24.38=(-*?=1*8=47(-.)S*8=9*77*897*8=2.=49745-*8$= *73&7)`=+3*3=b=11*2*398`=+322=b=27*(.:=]=1:77&-`=+323=b=-*++*7843= *9=&1_` =,**/'=b=,**1_= 1*8=47(-.)S*8=2.=49745-*8$=8439=&884(.S*8=&:88.=&;*(=)*8=(-&25.,3438=*(942>(47-.?.*38`= 85S(.+.6:*8=4:=343=5*&73&1*>`=,**1=b=.&'1*&:=._= =

24 3(&)7S=,_= &='.414,.*=)*8=47(-.)S*8=7S8:2S*=*3=(.36=54.398=*88*39.*18 =a= 8 :3*= )*8= +&2.11*8= )&3,.485*72*8= 1*8= 51:8= ).;*78.+.S*=&;*(= 57T8= )*= -*=***= *85T(*8= 17.''`= ,**-= 8 1*8=,7&.3*8=1*8=51:8=5*9.9*8=)*8=&3,.485*72*8=/*8+**= 2`=7).99.`=,***= 8 )*8= ,7&.3*8= )S5*3)&39*8= )*= (-&25.,3438= 54:7= 1*:7= ,*72.3&9.43= *9= 1*:7= 3:97.9.43= 11*2*398`=+322_==1B,*=&):19*`=1&=51:5&79=8439=*3=8>2'.48*=&;*(=(*8=(-&25.,3438=*9=1*:7= 574(:7*39=)*8=8:(7*8`=9&3).8=6:*=(*79&.3*8=*85T(*8=7*89*39=2>(4-S9S749745-*8=B.,_=/= 8 1&=51:5&79=8439=7&7*8=*9=).85*78*39=A=143,:*=).89&3(*=7).99.`=,***= 8 '&=51:5&79=439=)*8=8>89T2*8=)*=5411.3.8&9.43=97T8=85S(.&1.8S8=*3?.3,`=+32+= = B43(9.433*2*39= )*8= 7*1&9.438= 47(-.)S*88(-&25.,3438= &:= (4:78= )*= 1&= ;.*= )*= 147(-.)S*`= )*= 1&= ,*72.3&9.43=&=A=1B,*=&):19*='_='*=)S;*1455*2*39=):=57494(472*=*89=&88*?=1*39=0=24.8=A=/=&38`= $&82:88*3`=+33/=b=5-494=._=#&1.34;&_='=8>89T2*=7&(.3&.7*=)*= *5-&1&39-*7&=)&2&843.:2 =5-494=#_= $4>`=(4:5*=)*=7&(.3*=*9=2.(748(45.*=S1*(9743.6:*=A='&1&>&,*=)*=7&(.3*8=A=5*149438=*>71*`=+33/_= = &_=*72.3&9.43= 7494(472* =a=57*2.*78=89&)*8=)*=1&=,*72.3&9.43 = .4:9*8=1*8= 7&.3* = 9*258 = 47(-.)S*8=8439= )S5*3)&39*8= /*= 2 = 5*14943 8= ):3=(-&25.,343= 54:7=1*;*7=1&= )472&3(*=*9= 5*72*997*=1*:7= ,*72.3&9.43`=(&7= *2'7>43 = (*8=,7&.3*8=8439= 1&7'43*=47,&3.6:* = )S54:7;:*8=)*= 7S8*7;*8_== 47(-.)S* = -&25.,343=)=85S(.+.6:* = !&:`=8*18=2.3S7&:=`=&?49*=*9=5-485-47* =

'_=,*=&):19*= (74.88&3(* = .4:9*8=1*8= 47(-.)S*8= 9*77*897*8=8439= &884(.S*8=A=)*8= +*:.11*8=;*79*8 = (-&25.,3438=A= 1B,*=&):19*_='&= 51:5&79=)*8= 47(-.)S*8= (-14745->11.*33*8`= 5*14943 +4:73.88*39=)*8= 7&(.3* 8:(7*8=&:= (-&25.,343_= = 1=3*=8&,.9=5&8= 3S(*88&.7*2*39=):= 4:=)*8=2U2*8= /=(2 = /*= 2 = (-&25.,3438= 6:A=1&= 1&7'43*=47,&3.6:* = 47(-.)S* = -&25.,343=)=85S(.+.6:*= ,*72.3&9.43= !&:`=8*18=2.3S7&:=`=&?49*=*9=5-485-47*= = = =

25 .4:9*8= (*8= *85T(*8= 2>(4-S9S749745-*8= *9= 1*8= 6:*16:*8= *85T(*8= (-14745->11.*33*8= &884(.S*8=A=)*8=(-&25.,3438=*(942>(47-.?.*38=8439=)43(=.251.6:S*8=)&38=:3*=7*1&9.43=A= 974.8=5&79*3&.7*8=a=1*8=47(-.)S*8`=1*8=(-&25.,3438`=*9=1*8=&7'7*8=&884(.S8=8>2'.49.6:*2*39=A= (*8= 2U2*8= (-&25.,3438_= 1*8= 47(-.)S*8= 574+.9*39= )&38= (*= (&8= )*= 1&= (&5&(.9S= )*8= (-&25.,3438= *(942>(47-.?.*38= A= +&.7*= (.7(:1*7= ):= (&7'43*= 47,&3.6:*= )*5:.8= 1*(942>(47-.?*$=;*78=1*=2>(S1.:2=*3=)*-478=)*=1&=7&(.3*_= = *37*= -&25.,343= (414,.*= 5S(.+.(.9S= &'.9&9= S+S7*3(*= ) 7(-.)S*= 2>(47-.?.*3= :949745-*8======.5&7.8 = ":1&83*11&= +&.'1*= 9*77*897*= #(1472.(0= *9=&1_` =,**.= >57.5*).:2 == ":1&83*11&= +479*= 9*77*897*= -*++*7843= *9=&1_` =,**1= = "41:23.&= *7&94'&8.).:2= +&.'1*= *5.5->9*= %9*74= *9=&1_` =,**,= !43458.8== *7&94'&8.).:2= +479*= *5.5->9*= %9*74= *9=&1_` =,**.======.=49745-*8======9*7489>1.8= $-.?4(943.&= +479*= 9*77*897*= 7<.3= *9=&1_` =,**1= *5-&1&39-*7&= .-S1S5-47&(S*8=!1#= +&.'1*= 9*77*897*= <:14:= *9=&1_` =,**/= 5.5&(9.8= 8(42>(T9*8== +&.'1*= 9*77*897*= *1488*= *9=&1_` =,**.'= .24)47:2= $:88:1&(S*8= +479*= 9*77*897*= .71&3)&= *9=&1_` =,**0= = &'1*&:=._= I:*16:*8=*=*251*8=)*=1&=).;*78.9S=)*8=(-&25.,3438=2>(47-.?.*38=)47(-.)S*8=;*79*8=*9= )*= 1*:7= 85S(.+.(.9S_= 03*= 85S(.+.(.9S= +479*= 8.,3.+.*= 6:*= 94:9*8= 1*8= 47(-.)S*8= )*= (*99*= *85T(*= 8439= &884(.S*8=A=)*8=(-&25.,3438=):=2U2*=(1&)*`=6:*11*=6:*=84.9=1&=?43*=)S9:)*_=.=1*8=(-&25.,3438= &884(.S8=A=).++S7*39*8=47(-.)S*8=)*=1&=2U2*=*85T(*=&55&79.*33*39=A=)*8=(1&)*8=).++S7*398`=).89&398= 5->14,S3S9.6:*2*39`=1&=85S(.+.(.9S=)*=(*99*=&884(.&9.43=*89=+&.'1*_= = = 3(&)7S=-_= *8=7-.?4(943.&8$= 1*= 8439= )*8= (-&25.,3438= '&8.).42>(T9*8= 6:.= &55&79.*33*39= A= )*8= ,74:5*8= ).89&398= 5->14,S3S9.6:*2*39_=18=439=S9S=7*,74:5S8=84:8=1*=9*72*==7-.?4(943.&$=):=+&.9=)*=1*:78= 7*88*2'1&3(*8= 2475-414,.6:*8`= *3= (:19:7*= 4:= )&38= 1*8= (*11:1*8= )*8= 47(-.)S*8_= 1*8= (-&25.,3438= 8439= )*8= 5&7&8.9*8= 7&(.3&.7*8= 4:= )*8= 8&5745->9*8= )*8= 8418= &>2&3= *9= &1_` = ,**2_= I:&97*= 57.3(.5&:== ,*37*8= +472*39= )*8= 2>(47-.?*8= (-*?= 1*8= 47(-.)S*8=a= *7&94'&8.).:2`= ":1&83*11&`= "-&3&9*5-47:8 = *9= #*'&(.3& = F&7(:5`= +301=b= #447*`= +321=b= $&82:88*3`=+33/_=03*=2U2*=*85T(*=)*=7-.?4(943.&=5*:9=U97*=A=1&=+4.8=5&7&8.9*=*9=+472*7= )*8=2>(47-.?*8=&;*(=)*8=47(-.)S*8=#447*`=+321_== = 1*79&.38= *7&94'&8.).:2 = *9= #*'&(.3& = 8439= .)*39.+.S8= (422*= *(942>(47-.?.*38= 1*3= *9= &1_`= ,**,=b= *1488*= *9= &1_` = ,**,&=b= F*.88= *9= &1_` = ,**.=b= *1488*= *9= &1_` = ,**1=b= &&,&2*= *9= &1_` = ,**2'_= 1-*?=1*8=2>(4-S9S749745-*8`=43=7*974:;*=)*8=7-.?4(943.&8= (+_ &79.(1*= `=*88*39.*11*2*39= )*8= 7-.?4(943.&8= *(942>(47-.?.*38`= 6:.= ).++T7*39= )*8= 7-.?4(943.&8= &884(.S8= &:== 47(-.)S*8= ;*79*8=)*=5&7=1*:7=S(414,.*=*9=&:88.=)*=5&7=1*:7=548.9.43=5->14,S3S9.6:*_=&7=*=*251*`=1*8= *'&(.3&1*8=8439=8(.3)S*8=*3=)*:==(1&)*8`=).++S7*398=5&7=1*:7=S(414,.*=a=:3=(1&)*=7*,74:5*= )*8= 7-.?4(943.&8= .)*39.+.S*8= (-*?= )*8= 47(-.)S*8= ;*79*8= *9= 1&:97*= (1&)*= (477*8543)= &:== 8*'&(.3&1*8= *(942>(47-.?.*33*8= &884(.S*8= &:88.= A= )*8= 2>(4-S9S749745-*8= F*.88= *9= &1_` = ,**._=

26 3(&)7S=._= *8=7S8*&:==*(942>(47-.?.*38=(422:38= = *=1*(942>(47-.?*c= c&:=8>89T2*=7&(.3&.7*_= 2>(S1.:2 = = = '*8=*(942>(47-.?*8=!1#=8439= 1422*=1*8=*(942>(47-.?*8= !1# = )*8=47,&3*8=2.=9*8=+472S8=5&7= 8439=.3).;.):&1.8S*8=*9=6:*=1*= )*8=(*11:1*8=7&(.3&.7*8=*3= 8>89T2*=7&(.3&.7*=)*8=&7'7*8=*89= 8>2'.48*=&;*(=1*8=->5-*8=):3= 97T8=S9*3):`=:3=2U2*=&7'7*= (-&25.,343`=41=1*=(-&25.,343= 5*:9=U97*=&884(.S=A=51:8.*:78= *85T(*8=)*=(-&25.,3438=*3= ,=22 = 7&(.3* = 7*O4.9=)*8=8:(7*8=*9=+4:73.9=)*= 1*&:=*9=)*8=8*18=2.3S7&:== 2U2*=9*258=51:8.*:78= *88*39.*11*2*39=&?49*=*9= (*39&.3*8=)*85T(*8=5&7+4.8_= 5-485-47*=A=1&=51&39*_= = :=8>89T2*=7&(.3&.7*c= c&:=2>(S1.:2_= 1&39:1*=)*=.3 = = = '*=2>(S1.:2=):=(-&25.,343`= '*=2>(S1.:2=):=(-&25.,343= ).++:8=)&38=1*=841`=8S9*3)= 5*:9=+472*7=)*8= /:86:A=51:8.*:78=2T97*8=&:8 *(942>(47-.?*8=&;*(=51:8.*:78= )*1A=)*=1*(942>(47-.?*=*9= &7'7*8=).++S7*398`=)*=1&=2U2*= &'847'*=*&:=*9=3:97.2*398=)&38= *85T(*=4:=)*85T(*8=).++S7*39*8`= :3=51:8=,7&3)=;41:2*=*9=&;*(= )&38=:3=7&>43=)*=51:8.*:78= :3*=51:8=,7&3)*=*++.(&(.9S=6:*= 2T97*8_= 3*=1*=+*7&.*39=)*8=54.18= !3=(438S6:*3(*`=(*8= &'847'&398_= &884(.&9.438=2:19.51*8=5*:;*39= = ).34(:1*7=)*8=51&39:1*8 `=4:=)*= = +&(.1.9&9*7=1.389&11&9.43 =):3*= 2>(S1.:2 = &:97*=*85T(*=&884(.S*=&:== 2U2*8=(-&25.,3438= = *=1*(942>(47-.?*=A=)&:97*8=*(942>(47-.?*8_= = '*=2>(S1.:2=):=(-&25.,343=5*:9=97&38+S7*7=1*= (&7'43*=47,&3.6:*=6:.1=7*O4.9=&:=3.;*&:=):3*= *(942>(47-.?*=A=94:9=843=2>(S1.:2`=;4.7*=A= )&:97*8=*(942>(47-.?*8=6:.1=+472*=&;*(=:3= &:97*=-9*_=5&38=(*=(&8`=1*=(-&25.,343=5&79&,S= 5&7=1*8=)*:==-9*8=5*72*9=)*8=97&38+*798=)*= (&7'43*=*397*=&7'7*8=)*=1&=2U2*=*85T(*=4:= )*85T(*8=).++S7*39*8_= = 5*=9*18=+1:==)*=(&7'43*=5*:;*39= +&(.1.9*7 = 1S9&'1.88*2*39=)*8=51&39:1*8=4:=):3*=&:97*= *85T(*=*9=/4:*7=:3=71*=.25479&39=)&38=1*= +43(9.433*2*39=)*8=S(48>89T2*8 `=1&= (425S9.9.43= *397*=1*8=*85T(*8=;S,S9&1*8=*9=1&= )>3&2.6:*=)*8=(422:3&:9S8_ = B1:==)*&:=*9=)*=8*18=2.3S7&:== B1:==)*=(&7'43*=47,&3.6:*=*3=574;*3&3(*=):3=&7'7*= B1:==)*=(&7'43*=47,&3.6:*=*3=574;*3&3(*=):3=(-&25.,343= =

27 +_0 '*8=7S8*&:==2>(47-.?.*38=(422:38=1#8= = '*8= (-&25.,3438= 2>(47-.?.*38= +472*39= )*8= 2>(47-.?*8= &;*(= 1*8= 7&(.3*8= )*8= 51&39*8`=*9=7*O4.;*39=)*8=8:(7*8=6:.=1*:78=5*72*99*39=)S9*3)7*=1*:7=2>(S1.:2='.*3=&:8)*1A= )*= 1&= 7-.?485-T7*_= '*= 2>(S1.:2= ):3= 2U2*= .3).;.):= (-&25.,343= 5*:9= &.38.= *397*7= *3= (439&(9=&;*(=51:8.*:78=7&(.3*8=):3*=2U2*=51&39*`=2&.8=&:88.=&;*(=1*8=7&(.3*8=):3*=&:97*= 51&39*= )*= 1&= 2U2*= *85T(*`= ;4.7*= ):3*= *85T(*= ).++S7*39*`= 8*143= 843= )*,7S= )*= 85S(.+.(.9S= .&'1*&:= +`= *3(&)7S= .= *9= 1S9*3):*= )*= 843= 2>(S1.:2_= '&= 8>2'.48*= &;*(= 1*8= ,142S742>(T9*8= *89= *3= ,S3S7&1= 5*:= 85S(.+.6:*= *9= 5*:= S9*3):*_= '*8= '&8.).42>(T9*8= *9= &8(42>(T9*8= *(942>(47-.?.*38= 8439= 5&7+4.8= :3= 5*:= 51:8= 85S(.&1.8S8= #41.3&= *9= &1_` = +33,`= 2&.8=1&=51:5&79=8439=84:;*39=,S3S7&1.89*8= 47943=*9=7:38`=+332=b=,**+=b=7:38= *9=&1_` =,**,= *9=1*=2>(S1.:2=)*=(*8=(-&25.,3438=*89='*&:(4:5=51:8=S9*3):=)&38=1*=841=51:8.*:78=2T97*8= 4:= ).?&.3*8= )*= 2T97*8`= 2.9-= ]= $*&)`= ,**2_= .38.`= )*= +,= A= 3*= )*8= (-&25.,3438= *(942>(47-.?.*38=):3*=(422:3&:9S=8439=*3=,S3S7&1=&884(.S8=A=51:8.*:78=-9*8=A*33*)>= *9=&1_` =,**-=b=$.(-&7)= *9=&1_` =,**/=b=.<.*,= *9=&1_` =,**1_=!3=(438S6:*3(*`=1*8=*85T(*8=;S,S9&1*8= &884(.S*8=&:==2U2*8=*85T(*8=)*=(-&25.,3438=8439=549*39.*11*2*39=7*1.S*8=5&7=:3=7S8*&:= 2>(S1.*3= (422:3`= *3= &'7S,S= 1#8= =142243= #>(477-.?&1= 8*9<470==b= .2&7)= *9= &1_` = +331`=*1488*= *9=&1_ `=,**0_== = '&= 5488.'.1.9S= )*= 97&38+*798= )*= 3:97.2*398= *397*= ).++S7*39*8= *85T(*8= ;S,S9&1*8= A= 97&;*78= (*= 1.*3= 5->8.6:*= &= S9S= )S(4:;*79= .1= >= &= :3*= ).?&.3*= )&33S*8= )&'47)= (-*?= )*8= *85T(*8= *3)42>(47-.?S*8= *3= 1&'47&94.7*= 5:.8= (-*?= )*8= *85T(*8= *(942>(47-.?S*8= .3= 8.9: `= *3(&)7S=.=*9=843=S9*3):*`=*9=8&=51&(*=)&38=1*=+43(9.433*2*39=)*8=S(48>89T2*8`=&=84:1*;S= )*=342'7*:8*8=6:*89.438=*9=)S'&98=*1488*= *9=&1 _`=,**0_=.*3=6:*=1*=+43(9.433*2*39=)*=(*8= 1#8=&.9=S9S=5*:=&3&1>8S= .3=8.9: `=*9=8*2'1*=;&7.*7=&:=(4:78=)*=1&33S*='*7&9= *9=&1 _`=,**-`=*9= 8*143=1*8='*84.38=9745-.6:*8=)*8=51&39*8=.2&7)= *9=&1_` =+331`=(*8=1#8=54:77&.*39=/4:*7=:3= 71*=(*397&1=)&38=1*=+43(9.433*2*39=)*8=S(48>89T2*8=+1:==)*=(&7'43*=)&38=1*=841`=.*)*7844= *9=&1_` =,**1&`=1&=)>3&2.6:*=)*8=(422:3&:9S8=+&(.1.9&9.43=)*=1S9&'1.88*2*39=)*=(*79&.3*8= *85T(*8`= 5&7= 1&= 97&382.88.43= ).34(:1:2= )*= (-&25.,3438= 2>(47-.?.*38= (422:38= *9= 1&55479= )*= (&7'43*= &:== 51&39:1*8`= 4:= A= 1.3;*78*= (425S9.9.43=b= *;*7`= ,**,=b= %3,:*3*=]= A:>5*7`=,**,=*9=54:7=1&='.4).;*78.9S=2&.39.*3=)*=51:8.*:78=*85T(*8=6:.=&55&79.*33*39=&:= 2U2*= 7S8*&:=b= '*&0*= *9= &1_` = ,**.=b= .2&7)= ]= 5:7&11`= ,**.=b= #( :.7*`= ,**1_= 5*= 51:8`= 1*8= ).++S7*39*8=*85T(*8=)*=51&39*8=5&79&,*&39=)*8=2>(47-.?*8=5*:;*39=&:88.=+4:73.7=1.34(:1:2= 54:7= 1*8= 51&39:1*8= 343= 8*:1*2*39= )*= 1&= 2U2*= 2&.8= &:88.= )&:97*8= *85T(*8= 47943= *9= 7:38`=,**+=b=5.(0.*= *9=&1_` =,**.=b=$.(-&7)= *9=&1_` =,**/`=+&(.1.9&39=&.38.=1S9&'1.88*2*39=)*=(*8= *85T(*8=5&7=1&=97&382.88.43=).34(:1:2=$.(-&7)= *9=&1_`= ,**3_= =

28 &= 7'7*=-=-9*= = &5.3=5*(9.3S= U97*= 1-B9&.,3.*7= 1-&72*=*9=1-U3*=8*88.1*= 1-U3*=8*88.1*= U97*= *9= 1-U3* =8*88.1* = +=(2 = 77> = '= ;*11U2* = +/*02= B89-*`*,3 =

.,*02= B89-*`*+2 = ./*02= B89-*`*, = .1&9 =

9*=& = 9*=& =

B89 =-=*`*+/ =

B89 =-=*`*,1 = 9*=' = 9*=' =

B89 =-=*`*,3 = $S,.43= = $S,.43= =

= B.,:7*=/_ =1&7545-47*8=)*= &((&7.&=&2*9->89.3& =&`=51&3=)S(-&39.11433&,*=*3=B7&3(*=*9=.3).(*8=)*= ).++S7*3(.&9.43= ,S3S9.6:*= B 89 `= F7.,-9`= +30/= *397*= 7S,.438= '= 4:= *397*= 545:1&9.438= (_= '*8= 545:1&9.438= (477*8543)*39= A= )*8= ,74:5*8= ).3).;.):8= S(-&39.11433S8= 8:7= :3*= 2U2*= 5&7(*11*= (&7&(9S7.8S*=5&7=1*85T(*=)&7'7*=)42.3&39*=--9*_='*8=5&7(*11*8=8439=24348=4:='.885S(.+.6:*8_= =

29 +_1 '*8=1#8=)&38=1-&'.9&9=)*8=2>(4-S9S749745-*8= = '*8=51&39*8=2>(4-S9S749745-*8=8439=54:7=1&=51:5&79=&884(.S*8=A=)*8=(-&25.,3438= 2>(47-.?.*38`=(*=6:.=574:;*=)43(=6:*=(*:=8(.=8439=A=1&=+4.8=(&5&'1*8=)*=8&884(.*7=&;*(=:3= &7'7*= *9= :3*= 51&39*= 2>(4-S9S749745-*`= 84.9= )*:== *85T(*8= *9= )*:== 2475-4,S3T8*8= 97T8= ).++S7*39*8_=1*8=51&39*8`=*3=:9.1.8&39=)*=1&=2&9.T7*=47,&3.6:*=574):.9*=5&7=)*8=&:949745-*8`= *9=97&38+S7S*=5&7=)*8=(-&25.,3438`=)S24397*39=.3).7*(9*2*39=1*=.89*3(*=)*=1#8=*1488*= *9= &1_` = ,**0_= 1*5*3)&39`= (*8= 51&39*8= 2>(4-S9S749745-*8= )S94:73*39= 1*8= (-&25.,3438= *(942>(47-.?.*38= )*= 1*:7= +43(9.433*2*39= 3472&1=b= .1= *89= )43(= ).++.(.1*= )*3= )S):.7*= )*8= ,S3S7&1.9S8= *9= 349&22*39= )*= 8&;4.7= 8.= 51:8.*:78= &:949745-*8= 7*1.S8= 5&7= (*8= 2U2*8= (-&25.,3438=+43(9.433*7&.*39=)*=1&=2U2*=+&O43_= 'S9:)*= (425&7&9.;*= )*8= (422:3&:9S8= )*= (-&25.,3438= 57S8*398= 8:7= 1*8= 7&(.3*8= )&7'7*8=574(-*8=8:,,T7*=6:*=1&=2U2*=*85T(*=)*=(-&25.,343=(4143.8*=*++*(9.;*2*39=)*8= &7'7*8=)*85T(*8=).++S7*39*8`=*9=)*=342'7*:8*8=*85T(*8=)*=(-&25.,3438=*(942>(47-.?.*38= &55&7&.88*39=*++*(9.;*2*39=(422*=,S3S7&1.89*8= 47943=]=7:38`=+332=b=,**+_=1*5*3)&39`= .1= 8&,.9= 84:;*39= ):3*= .)*39.+.(&9.43= 2475-414,.6:*= )*8= *(942>(47-.?*8= 4:= )*8= (&7545-47*8`=4:=):3*=.)*39.+.(&9.43=A=1&.)*=)*8=._=%7=(*8=)*:==&5574(-*8=3*=5*72*99*39= 5&8= )*= )S9*(9*7= )*8= (&8= )*= 85S(.&9.43= (7>59.6:*= (43):.8&39= A= )*8= 7&(*88-9*8`= :3= 5-S342T3*= 8&38= )4:9*= +7S6:*39= (-*?= 1*8= (-&25.,3438= *(942>(47-.?.*38= &3*3= *9= &1_` = ,***=b=)*3=&00*7= *9=&1_` =,**.=b=54:-&3= *9=&1_` =,**/_=+.3=)*=)S9*72.3*7=8.=1*=2U2*=.3).;.):= 5*:9=(4143.8*7=*++*(9.;*2*39=51:8.*:78=*85T(*8=).++S7*39*8`=.1=+&:9=54:;4.7=).++S7*3(.*7=1*8= .3).;.):8`= *9= 1*8= .= 4:= 1&= 2475-414,.*= 3*= 8439= 51:8= .(.= )&:(:3*= :9.1.9S_= !3= 7*;&3(-*`= 1:9.1.8&9.43= )*= 2&76:*:78= 241S(:1&.7*8= (4:7&22*39= :9.1.8S8= *3= ,S3S9.6:*= )*8= 545:1&9.438`= 9*18= 6:*= )*8= 2.(748&9*11.9*8`= 5*72*9= 343= 8*:1*2*39= )*= ).++S7*3(.*7= )*8= .3).;.):8=2&.8=&:88.=)*=)S9*72.3*7=1&=897:(9:7*=)*8=545:1&9.438=54:7=*=(1:7*=1*=.89*3(*= )*85T(*8='.414,.6:*8=(7>59.6:*8_= 5&38=1&=2*8:7*=41=1S9:)*=)*=1&=2>(4-S9S749745-.*=7*548*=8:7=1*=.89*3(*=)*=1#8`= /&.= (422*3(S= 5&7= )S24397*7= 1&'8*3(*= )*= 85S(.+.(.9S= )*= (-&25.,3438= *(942>(47-.?.*38= (+_ &79.(1*= _='*8=+1:==)*=,T3*8=*397*=51:8.*:78=545:1&9.438=)*=&((&7.&=&2*9->89.3&= B.,_=/&= 54:88&39= 84:8= )*8= *85T(*8= )&7'7*8= ).++S7*39*8= 439= S9S= S9:).S8= B.,_= /'= A= 1&.)*= )*= 2&76:*:78= 2.(748&9*11.9*8= *9= )*= 2&76:*:78= ):= 541>2475-.82*= )*= 1&= 143,:*:7= )*= +7&,2*398= &251.+.S8= 5'`= 5*82&7&.8= *9= &1_` = +332_= 1*= (-&25.,343= &55&7*22*39= ,S3S7&1.89*= 3*89= 5&8= &884(.S= &:== 51&39*8= 2>(4-S9S749745-*8`= 5&7= (4397*`= .1= 8&,.9= ):3= (-&25.,343= )*= +47U98= 51:99= B,S*8`= (422*= (*11*8= 41= 8*= 974:;*39= 84:;*39= 1*8= *85T(*8= 2>(4-S9S749745-*8_=1=7*88479=6:*= &((&7.&=&2*9->89.3& =3*=(&(-*=5&8=)*=7&(*=85S(.&1.8&9.43= (7>59.6:*`=*9=6:*=1*8=+1:==)*=,T3*8=*397*=545:1&9.438=2U2*=).89&39*8=)*=/**=02=*9=84:8= )*8=&7'7*8=).++S7*398=8439=&88*?=.25479&398=B.,_=/(_=1*(.=)S24397*=343=8*:1*2*39=6::3*= 2U2*=*85T(*=*89=&884(.S*=A=51:8.*:78=&7'7*8=).++S7*398=2&.8=&:88.`=6:*=1&884(.&9.43=3*=9*3)= 5&8=3S(S88&.7*2*39=A=8*=85S(.&1.8*7_=&38=)S24397*7=1*=+43(9.433*2*39=)*=1#8=)&38=(*8= +47U98`= (*8= 57*2.T7*8= 57*:;*8= ,S3S9.6:*8= .3= 8.9: = )*= 1&'8*3(*= )*= 85S(.&9.43= ;.88A8;.8= )*= 1-9*`=24397*39=6:*=1*8=1#8=5*:;*39=8*=+472*7=*397*=).++S7*39*8=*85T(*8=;S,S9&1*8_= = 1*5*3)&39`=(422*=1*=).9=54'?-&380.=+31-`== 7.*3=3&=)*=8*38=*3='.414,.*=8.1=(*=3*89= A= 1&= 1:2.T7*= )*= 1S;41:9.43= _= .*3= 6:*= (*8= *85T(*8= A= 1&= ;.*= 84:9*77&.3*= 2>89S7.*:8*= 3*= (4389.9:*39=6::3*=(:7.48.9S=):=243)*=;S,S9&1=54:7=1".1=):=3&9:7&1.89*`=*11*8=84:1T;*39=)*= 342'7*:8*8=6:*89.438=6:&39=A=1&55&7.9.43=):3=9*1=24)*=)*=;.*`=*9=6:&39=A=843=S;41:9.43_==

30 =

4:8 897.':8= = -&25.,3438== (414,.* = &'.9&9 = 2>(47-.?.*38=

9&3-45*.3&* .*1.458.).3&* &=.11&7..3&* E>,45*9&1.3&* 3(.)..3&* 7.458.).3&* &9&8*9.3&* ;742-*&)..3&* :145-..3&* 1457.3&(S*8 = = .745.(&1 = >2'.)..3&* 3,7&*(.3&* *7&3,.).&* *7.).3&* $:88:1&(S*8 = = .745.( &1 = 41>89&(->.3&* ,748945->11.3&* 43*7.3&* ;1*9..3&* *'&(.3&1*8 = = .*25S7S = 1*:749-&11.).3&* &*1..3&* 4)4(-.1.3&* 411&'..3&* 4*14,>3.3&* 7*9-:8.3&* 7.5-47&* *7;.1.*&* 1457.3&(S*8 = = .745.(&1 = !5.)*3)74.)S*8 = &8974).*&* &5745->9*8 = `= = .745.(&1 = &1>584*&* >(45*7)43 = = .745.(&1 = &1&=.)*&* "-&3&9*5-47:8 = = .74 5.(&1 = *3)74'..3&* $:88:1&(S*8 = = .*25S7S = 745.).*& $:88:1&(S*8 = 4'7&1.*&* = .*25S7S = *499.*&* .-S1S5-47&(S*8 = = .*25S7S = -.?&39-*11.3&* *'&(.3&1*8= = .*25S7S = (.&39-.3&* &1&)*3..3&* = = 7>59489>1.).3&* .:7.).3&* "-&3 &9*5-47:8 = = .745.(&1 = %7(-.)4.)S*8 = 7&845->11.3&* 44)>*7.3&* *7&94'&8.).:2 = = .745.(&1 = &33.*11.3&* 7&3.(-.).3&* 5.7&39-.3&* 9*7489>1.).3&* -147&*.3&* 7(-.).3&* .8.3&* ;74<31**.3&* 4)4347(-.)*&* >57.5*).4.)*&* >45->11:2 = = .745.(&1 = C&3.11.4.)S*8 = &3.11.3&* &5745->9*8 = = .745.(&1 = 4,43..3&* 72.11&7.& = = .745.(&1 = 5489&84.)*&* = = = B.,:7*=0_= 11&88.+.(&9.43=5->14,S3S9.6:*=)*8=84:8897.':8=)%7(-.)S*8_='*8='7&3(-*8=(4147.S*8=*3=748*= (425479*39= )*8= *85T(*8= 2>(4-S9S749745-*8_= 'S(414,.*= )*8= (-&25.,3438= *89= )S):.9*= )*= 1*:7= .)*39.+.(&9.43=9&=4342.6:*=*(942>(47-.?.*3=a=!`=5&7&8.9*=a=`=8&5745->9*=a=_='*8=97.':8=*3=47&3,*8= 8439=(*11*8=S9:).S*8=&:=(4:78=)*=2&=9-T8*_=->14,S3.*=)&57T8=1-&8*= *9=&1_` =,**-_=

31 _ !;41:9.43= )*= 1&= 2>(4-S9S749745-.*= *9= ).;*78.9S=S(414,.6:*=)*8=2>(4-S9S749745-*8= = = ,_+ 'S;41:9.43=)*=1&=2>(4-S9S749745-.*=*3=2.1.*:=9*25S7S= = = 842'7*:== 8439= 1*8= &79.(1*8= 6:.= 97&.9*39= )*= 1S;41:9.43= )*= 1&= 2>(4-S9S749745-.*`= 349&22*39= (-*?= 1*8= 243497454.)S*8= *9= 1*8= 47(-.)S*8= *3= 8*= +43)&39= 8:7= 1*8= *85T(*8= 2>(4-S9S749745-*8= )S/A= S9:).S*8= *=_= .)&7943)4`= ,**/_= %7`= 94:9*8= 1*8= *85T(*8= 2>(4-S9S749745-*8=)*=(*8=)*:==+&2.11*8=S9:).S*8=/:86:A=57S8*39=*3=2.1.*:=9*25S7S=8439= &884(.S*8=97T8=85S(.+.6:*2*39=A=)*8=(-&25.,3438=*(942>(47-.?.*38=B.,_=0=b=.&>147= *9=&1_` = ,**,_= &(-&39= 6:*= 1*8= 47(-.)S*8= ;*79*8= 8439= &884(.S*8= A= )*8= (-&25.,3438= 8&5745->9*8= 7&88*2'1S8= 84:8= 1*= 9*72*= )*= 7-.?4(943.&$= $&82:88*3`= +33/`= )*= 2&3.T7*= 51:8= 4:= 24.38= 85S(.+.6:*`= 1*= 5&88&,*= A= 1&= 2>(4-S9S749745-.*= 8&((425&,3*= )43(= ):3= (-&3,*2*39= )*= 5&79*3&.7*= 2>(47-.?.*3= *9= )*8= S(-&3,*8= &;*(= (*= 5&79*3&.7*`= &.38.= 6:*= ):= 3.;*&:= )*= 85S(.+.(.9S=)*=(*99*=34:;*11*=&884(.&9.43=*1488*=*9=&1_` =,**0_= = '*8= 51:8= 574(-*8= 5&7*398= (-14745->11.*38= )*= $*499.&= 3.):88&;.8` =(422*= *5-&1&39-*7&= )&2&843.:2 `= _= 143,.+41.& = 4:= 5.5&(9.8= 2.(745->11&` = 8439= &:88.= &884(.S8= A= )*8= (-&25.,3438=*(942>(47-.?.*38`=349&22*39=)*8=9-S1S5-47&(S*8`=)*8=7:88:1&(S*8=2&.8=)*= 2&3.T7*=343=85S(.+.6:*=*1488*= *9=&1_` =,**.'=b=<:14:= *9=&1_` =,**/=b='&).*= *9=&1_` =,**0=9&3).8= 6:*= .24)47:2= &'479.;:2 = *89= &884(.S*= 85S(.+.6:*2*39= A= )*8= 7:88:1&(S*8= .71&3)&= *9= &1_` = ,**0_='&=8.,3&9:7*=.84945.6:*=)*=94:9*8=(*8=*85T(*8= +- 1=*9= +/ 8==*89=.39*72S).&.7*=*397*= (*11*=)*85T(*8=&:949745-*8=*9=(*11*=)*85T(*8=2>(4-S9S749745-*8=a=*11*8=7*O4.;*39=)43(=):= (&7'43*=)*=1*:78=(-&25.,3438=2>(47-.?.*38`='.*3=6:*11*8=84.*39=&:88.=5-4948>39-S9.6:*8= .)&7943)4= *9=&1_` =,**.=b=<:14:= *9=&1_` =,**/=b='&).*= *9=&1_` =,**0_=1*99*=8.9:&9.43=8:757*3&39*= 8*=51.6:*=5&7=1*=+&.9=6:*=1*8=47(-.)S*8=:9.1.8*39=)*:==9>5*8=)*=(&7'43*=a=.347,&3.6:* =;.&= 1&= 5-4948>39-T8*`= *9= 47,&3.6:* = ;.&= 1*:78= 2>(47-.?*8`= (*= 6:.= +&.9= )*11*8= )*8= 2.=49745-*8$= *1488*= *9=&1_` =,**0_=.38.`=1*=8*:1=97&.9=57457*=&:==*85T(*8=2>(4-S9S749745-*8=(477*8543)= A=1*:7=85S(.+.(.9S=)&884(.&9.43=B:72&3=]=.7&55*`=+31+=b=7:38= *9=&1_` =,**,=b=.&>147= *9=&1_` = ,**,_= = 1*5*3)&39`= (*99*= 85S(.+.(.9S= *89= 8:757*3&39*= )&38= 1&= 2*8:7*= 41= *11*= 1.2.9*= 1&= (&5&(.9S= )*= (*8= 47(-.)S*8= A= 974:;*7= 1*= '43= 5&79*3&.7*= 2>(47-.?.*3_= '*8= *=51.(&9.438= (422:3S2*39= &)2.8*8= &(9:*11*2*39= 7*548*39= 8:7= )*:== ->549-T8*8=a= 1&= (48&)&59&9.43= +43(9.433*11*= *9= 1&= (48*;41:9.43= 5&7&8.9&.7*= 7:38= *9= &1_` = ,**,_= '->549-T8*= )*= (48 &)&59&9.43= +43(9.433*11*= 89.5:1*= 6:*= 1*= 2S(&3.82*= 5&7= 1*6:*1= 1*8= 47(-.)S*8= 2>(4-S9S749745-*8= 7*O4.;*39= ):= (&7'43*= )*8= (-&25.,3438`= A= 1.3;*78*= )43(= )*8= +1:== -&'.9:*18=)&38=1*=(&)7*=):3*=8>2'.48*=1&2*743= *9=&1_` =,**2=3S(*88.9*=)*8=&)&59&9.438=8.= 342'7*:8*8=*9=85S(.&1.8S*8`=6:*=(*8=&)&59&9.438=3*=5*:;*39=U97*=)S;*1455S*8=;.88A8;.8=)*= 51:8.*:78= *85T(*8= )*= (-&25.,3438_= 1*5*3)&39`= (*1&= 3*= 5*72*9= 5&8= )*=51.6:*7= 1*8= +7S6:*398=(-&3,*2*398=)-9*8=4'8*7;S8=(-*?=1*8=#4349745*8=.)&7943)4`=,**/_= = =

32 3(&)7S=/_= *8=2>(4-S9S749745-*8=a=)*8=5&7&8.9*8== '*8=51&39*8=2>(4-S9S749745-*8=7*O4.;*39=1*:7=2&9.T7*=47,&3.6:*`=.3.9.&1*2*39=574):.9=5&7= :3= &:949745-*`= ;.&= 1*:78= (-&25.,3438= 2>(47-.?.*38_= !11*8= 3*= +4:73.88*39= &55&7*22*39= 5&8= )*= 2&9.T7*=47,&3.6:*=&:==(-&25.,3438=*9=8*=(425479*39=*3=5&7&8.9*8_='*:7=89&9:9=)*=5&7&8.9*`=8.1=*89= &((*59S=5&7=(*79&.38=.)&7943)4= *9=&1 _`=,**,=b='*&0*`=,**.=*89=3S&324.38=).8(:9&'1*= &7)*8`=,**,_= '*=5&7&8.9.82*=):3*=*85T(*==5&7=7&55479=A=:3*=&:97*=*85T(*==*89=)S+.3.=)&57T8=1&=(4397*8 8S1*(9.43= .3):.9*= (-*?= 1*85T(*= = 5&7= (*99*= .39*7&(9.43_= := 3.;*&:= .3).;.):*1`= 1.39*7&(9.43= *89= (429*:8*= 54:7= 1*8= )*:== *85T(*8`= 5:.86:*= 1*8= 2S(&3.82*8=)*=)S+*38*=)*=1*85T(*==5*:;*39=&:88.= S1.2.3*7=)*8=.3).;.):8=)*=1*85T(*=`=2&.8=51:8=,14'&1*2*39`=1*8=(4298=8439=51:8=.25479&398=54:7= 1*85T(*==6:*=54:7=1*85T(*=_='&=8>2'.48*=*397*=)*:== *85T(*8= 8*= 97&):.9= 5&7= :3*= 51:8= ,7&3)*= ;&1*:7=8S1*(9.;*=)*8=)*:==*85T(*8_=:=3.;*&:=.3).;.):*1`=1&=8>2'.48*=5*:9=&:88.=.3):.7*=)*8=(4298`= 2&.8= 6:.= 8439= (4397*'&1&3(S8= 5&7= )&:97*8= 'S3S+.(*8_= '&= ;&1*:7= )*8= (4298= )S5*3)= )*= 1*3;.7433*2*39`=*9=5&7=*=*251*=1*8=51&39*8=6:.=54:88*39=*3=2.1.*:=&?49S=3.97&9*8=1.2.9*39=1*:78= &884(.&9.438= 8>2'.49.6:*8= &;*(= 1*8= '&(9S7.*8= +.=&97.(*8= )&?49*`= (&7= 1&= 8>2'.48*= &;*(= 1*8= '&(9S7.*8= )*;.*39=51:8=(429*:8*=6:*=1&'84759.43=)*=3.97&9*8=97**9*7=]=F43,`=+322_=:=3.;*&:=.3).;.):*1`= (422*=1&=349.43=)*=(429=*89=7*1&9.;*`=1*=89&9:9=)*=1&884(.&9.43=5*:9=;&7.*7=8:7=:3=(439.3::2=*397*= 5&7&8.9.82*=*9=8>2'.48*_=5*=2U2*`=&:=3.;*&:=)*=1*85T(*`=1&=1.2.9*=*397*=8>2'.48*=*9=5&7&8.9.82*= )*;.*39= +14:*= *9= 1*8= &884(.&9.438= +472*39= &:88.= :3= (439.3::2_= 5&38= (*99*= 2*8:7*`= .1= 3*89= 5&8= 3S(S88&.7*=)*=97&3(-*7=*3=+&;*:7=):=89&9:9=5&7&8.9&.7*=4:=8>2'.49.6:*=)*8=2>(4-S9S749745-*8_= 5*= 51:8`= 5*:= )S1S2*398= 5*:;*39= 5*72*997*= )S9&'1.7= 1*8= (4298= )*= 1*:7= &884(.&9.43_= := 3.;*&:=)*=1*85T(*`=43=3*=8&.9=5&8=8.=1.39*7&(9.43=&;*(=1*8=2>(4-S9S749745-*8=(4397*88S1*(9.433*=1*8= *85T(*8=)*=(-&25.,3438=2>(47-.?.*38`=*9=1*8=&:949745-*8=6:.=>=8439=&884(.S8_=1=+&:)7&.9=54:;4.7= 8:.;7*= )*8= 545:1&9.438= 8:7= 1*= 143,= 9*72*`= 47= 1S;&1:&9.43= )*8= *++*(9.+8= )*= (-&25.,3438= *(942>(47-.?.*38= *89= )S1.(&9*= (&7= 1*8= .3).;.):8= 8439= ).++:8= )&38= 1*= 841`= *9= 1&'43)&3(*= )*8= (&7545-47*8=*3=7*57S8*39*=5&8=1&'43)&3(*=84:9*77&.3*=$.(-&7)= *9=&1 _`=,**/_=:=3.;*&:=.3).;.):*1`= .1=+&:)7&.9=54:;4.7=2*8:7*7=1*8=(4298=)*=(*99*=&884(.&9.43_=5:=54.39=)*=;:*=):=(&7'43*`=.1=+&:)7&.9= 54:;4.7=S;&1:*7=1*8=+1:==*9=343=5&8=8*='&8*7=8:7=1&= '.42&88*=(422*=(*8=51&39*8=7*85.7*39`=*11*8= 5*7)*39=(439.3:*11*2*39=:3*=5&79.*=):=(&7'43*=6:*11*8=7*O4.;*39`=(*=6:.=7*3)=1S;&1:&9.43=*3(47*= 51:8= ).++.(.1*_= '*8= (-&25.,3438= *(942>(47-.?.*38= ):3= &7'7*= 7*O4.;*39= *3;.743= +*= ):= (&7'43*= 574):.9= 5&7= (*9= &7'7*= .2&7)= ]= 5:7&11`= ,**._= 14259*= 9*3:= )*= 1&= ).++S7*3(*= )*= 9&.11*= *397*= 1&= '.42&88*= )*8= &7'7*8= *9= (*11*= )*8= 2>(4-S9S749745-*8`= .1= *89= 5488.'1*= 6:*= 1&= 5&79= ):= (&7'43*= 47,&3.6:*=7*O:=5&7=1*8=2>(4-S9S749745-*8=84.9=3S,1.,*&'1*=5&7=7&55479=A=1&=5&79=7*O:*=.3.9.&11*2*39= 5&7=1*8=(-&25.,3438=*(942>(47-.?.*38_=5*=51:8`=(*8=2>(4-S9S749745-*8=*3=57S1*;&39=:3*=5&79.*=):= (&7'43*= )*8= (-&25.,3438= 5*:;*39= +&.7*= )*:== :3= 51:8= ,7&3)= 5:.9= )*= (&7'43*`= *9= &:,2*39*7= 1*= 97&38+*79= )*= (&7'43*= 47,&3.6:*= )*= 1&:949745-*= ;*78= 1*8= (-&25.,3438= *++*9= 5:.9`= 54880*>= *9= &1_`= +33*=b= .)&7943)4= *9= &1_`= ,**+_= 5&38= (*= (&8`= 1&884(.&9.43= &;*(= 1*8= 2>(4-S9S749745-*= 54:77&.9= &:,2*39*7=1*8=+1:==)*=(&7'43*=;*78=1*8=(-&25.,3438`=*9=1*:7=&55479*7=)43(=)*8='S3S+.(*8=3:97.9.+8_= 5*=51:8`=.)&7943)4= *9=&1_ =,***=439=4'8*7;S=:3*=51:8=,7&3)*=(43(*397&9.43=)*8=2>(47-.?*8= )*= --.?454,43 =85_=A=574=.2.9S=)*= #&7(4)*8=8&3,:.3*& `=:3*=*7.(&(S*=2>(4-S9S749745-*8`=(*=6:.=24397*= 6:*= 1&= 57S8*3(*= )*= (*99*= 2>(4-S9S749745-*= 54:77&.9= 89.2:1*7= 1&= (74.88&3(*= ):= (-&25.,343_= 5*= 2U2*`= 1&= 57S8*3(*= )*= --.?454,43 = 85_= 89.2:1*= 1&= ,*72.3&9.43= )*= #&7(4)*8= 8&3,:.3*& `= 2U2*= 8&38= (439&(9= &;*(= 1*= (-&25.,343= 7:38= ]= $*&)`= ,***`= (*= 6:.= 8:,,T7*= 1&= 57S8*3(*= )-47243*8= )*= (74.88&3(*=4:=)*=8.,3&:==S(-&3,S8=*397*=1*8=)*:==5&79*3&.7*8_=1*8=8.,3&:==5*:;*39=&;4.7=:3=*++*9= 548.9.+=*3=9*72*8=)*=(74.88&3(*=2&.8=3*=8.,.3+.*=5&8=6:*=1&884(.&9.43=*89=8>2'.49.6:*`=(*8=8.,3&:== 5*:;*39=+&.7*=5&79.*=)*=1&=2&3.5:1&9.43=5&7&8.9&.7*=)*8=2>(4-S9S749745-*8_= 1*8=2>(4-S9S749745-*8=439='*&:=8*=(425479*7=*3=5&7&8.9*=*9=2&3.5:1*7=1*:78=(-&25.,3438= 2>(47-.?.*38`= 1*= (429= )*= (*8= 2>(4-S9S749745-*8= *89= 5745&'1*2*39= 3S,1.,*&'1*= 54:7= (*8= (-&25.,3438= 6:.= 7*O4.;*39= )*= ,7&3)*8= 6:&39.9S8= )*= (&7'43*= 47,&3.6:*= )*8= &7'7*8= *1488*= *9= &1 _`= ,**0_= !3(47*= :3*= +4.8`= (*99*= &884(.&9.43= 5*:9= U97*= (438.)S7S*= (422*= :3= (439.3::2`= 41= 1*8= 2>(4-S9S749745-*8=(4389.9:*39=:3=*=97U2*`=*9=41=1*8=51&39*8=2.=49745-*8$=6:.=7*O4.;*39=24.38=)*= (&7'43*=6:*=1*8=2>(4-S9S749745-*8`=7*57S8*39*39=:3=S9&9=.39*72S).&.7*_=

33 '&:97*=->549-T8*`=)*=(48*;41:9.43=5&7&8.9&.7*`=+&.9=.39*7;*3.7=1*=24)T1*=)*=1&=7*.3*=74:,*= C&3=C&1*3`=+311=b=11&>=]=A4;*7`=+330=8*143=1*6:*1`=)&38=:3*=&884(.&9.43=5&7&8.9&.7*`=1*8= )*:== 5&79*3&.7*8= 9*39*39= 1:3= )S(-&55*7= &:= 5&7&8.9.82*`= 1&:97*= )*= 2&.39*3.7= 1*= 5&7&8.9.82*_=1=8*38:.9=:3*=(4:78*=S;41:9.;*==&:==&72*2*398=`=6:.=8S1*(9.433*=:3*=51:8= ,7&3)*= 85S(.+.(.9S= ).39*7&(9.43= *3= 7S):.8&39= 574,7*88.;*2*39= 1*= 342'7*= ).39*7&(9.438= (425&9.'1*8_= '*= 89&9:9= 5&7&8.9&.7*= )*8= 2>(4-S9S749745-*8= +&.9= 94:/4:78= )S'&9= *3(&)7S= /= 2&.8=3*=1.2.9*=5&8=1*8=7S+1*=.438=8:7=1&=85S(.+.(.9S_='&=9-S47.*=)*=1&=7*.3*=74:,*=8&551.6:*= &:88.= &:== &884(.&9.438= 8>2'.49.6:*8`= )&38= 1&= 2*8:7*= 41= (*8= )*73.T7*8= 5*:;*39= &:88.= .3+1.,*7=)*8=(4298`=(*=6:.=8S1*(9.433*=(-*?=1*8=)*:==5&79*3&.7*8=)*8=&)&59&9.438=6:.=1.2.9*39= (*8= (4298`= *9= (43):.9= A= :3*= 51:8= ,7&3)*= 85S(.+.(.9S= )*= 1&884(.&9.43_= 5&38= (*= (&)7*`= 1*8= 51&39*8= 2>(4-S9S749745-*8= 8*= 85S(.&1.8*7&.*39= ;.88A8;.8= )*8= (-&25.,3438= 1*8= 24.38= (429*:== 6:.= &55479*39= 51:8= )*= (&7'43*= 5&7= *=*251*`= (*= 6:.= *=51.6:*7&.9= 6:*= )*8= 2>(4-S9S749745-*8= )*= +&2.11*8= ).++S7*39*8= 84.*39= &884(.S*8= &:== 2U2*8= (-&25.,3438`= (422*= 47&1147-.?&=2&(:1&9& =*9= 4349745& =:3.+147& `=94:9*8=)*:==&884(.S*8=A=)*8=7:88:1&(S*8`= )*8=(-&25.,3438=*(942>(47-.?.*38=+7S6:*398=*3=2.1.*:=9*25S7S=.&>147=]=7:38`=+333=b= .)&7943)4=*9=7:38`=,**,=b=.&>147= *9=&1_` =,**._='&=85S(.+.(.9S=)*8=2>(4-S9S749745-*8=&=S9S= .11:897S*=5&7=)*=342'7*:==*=*251*8`=(422*= 47&1147.?&=85 =.&>147=]=7:38`=+333=b=.&>147= *9= &1_` =,**,=4:=1*8=#4349745*8=)439=1*8=5&79*3&.7*8=8439=).++S7*398=*397*=*85T(*8=;4.7*=*397*= 545:1&9.438`= (*= 6:.= 24397*= &:88.= :3*= 9*3)&3(*= A= 1&= 8S1*(9.43= ):3*= 85S(.+.(.9S= 14(&1*= .)&7943)4=]=7:38`=,**+=b=,**,=b=.)&7943)4`=,**/_= 1*5*3)&39`=(*8=24)T1*8=)S;41:9.43=3*=8*=+43)*39=6:*=8:7=)*8=(&8=9*25S7S8`=41=1*8= +47U98=8439=)42.3S*8=5&7=)*8=&7'7*8=*(942>(47-.?.*38`=9&3).8=6:*=1&=2&/47.9S=)*8=47(-.)S*8= 2>(4-S9S749745-*8= *89= 9745.(&1*_= %7`= 1*8= 8>89T2*8= *(942>(47-.?.*38= 8439= 5*:= &'43)&398= 84:8= 1*8= 9745.6:*8`= *=(*59S= *3= 8.*`= *9= 1*8= 47(-.)S*8= 9745.(&1*8= 57S8*39*39= :3*= ,7&3)*= ).;*78.9S=)*=5&79*3&.7*8=2>(47-.?.*38=549*39.*18_=1*79&.3*8=8439=&884(.S*8=85S(.+.6:*2*39=A= )*8=(-&25.,3438=5&7&8.9*8`=9*11*=6:*= 72.11&7.&=2*11*&= .&'1*&:=-`=.*7&8-.9&`=+32/=b=%,:7&8 .8:/.9&= *9=&1_ =,**3`=4:=8&5745->9*8=9*1=6:*=)*8=1457.3&(S*8=(-*?=5.54,.:2=748*:2 =.&'1*&:= -`=&&2&94= *9=&1_` =,**/_= = 7(-.)S*8= 884(.&9.43=A== .1.*:= .1.*:= )*8=(-&25.,3438= *25S7S= 9745.(&1= >(4-S9S749745-*8= !(942>(47-.?.*38= (((= (= &5745->9*8= 8= (((= &7&8.9*8= 8= (((= 5S(.+.6:*8= (((= (((= .=49745-*8= 5S(.+.6:*8= (=4:=8= = !(942>(47-.?.*38= (((= = = &'1*&:= /_= 1425&7&.843= )*= 1&884(.&9.43= 2>(47-.?.*33*= )*8= 47(-.)S*8= 2>(4-S9S749745-*8= *9= 2.=49745-*8$=*3=2.1.*:=9*25S7S=*9=9745.(&1_= = 1*8= ).++S7*3(*8= &;*(= 1*8= 24)T1*8= 9*25S7S8= 54:77&.*39= U97*= &997.':S*8= &:= 2.1.*:= 9745.(&1=*9=548*39=974.8=6:*89.438=9&'1*&:=2=a== +_ 1&= ).;*78.9S=S(414,.6:*=)*8= (-&25.,3438= 2>(47-.?.*38= )47(-.)S*=*898*11*= 57457*= &:= 2.1.*:= 9745.(&1== 6:*11*= *89= 1&= 5&79= )*8= (4397&.39*8= 5->14,S3S9.6:*8= *9= S(414,.6:*8=8:7=(*8=&884(.&9.438=2>(47-.?.*333*8== ,_ 1&=85S(.+.(.9S=*898*11*=94:/4:78=)*=7.,:*:7=)&38=1*8=)*:==S(48>89T2*8== -_ 1&=2.=49745-.*=*=.89*89*11*=*3=2.1.*:=9745.(&1==

34 ======*499.S*8= *5-&1&39-*7& `= *7;.1. S*8= 5.54,.:2 = = &1>584S*8 `=;1*99.3S*8 ======`=`= `=`= `= = `=`= = = = = `=`=`= = = = = ======Nord = *499.S*8= *5-&1&39-*7& `= *7;.1.S*8= 5.54,.:2 = 1000 km = &1>584S*8`=:145-.S*8`=*7.).3S*8`=&8974).S*8= = -.?&39-*11.3S*8`=44)>*7.3S*8`=&3.11.3S*8`= 4,43.S*8======B.,:7*= 1_= $S5&79.9.43= )*8= 3*499.S*8= 1*997*8= 2&/:8(:1*8= *9= )*8= 84:8897.':8= )47(-.)S*8= 2>(4-S9S749745-*8= (&)7*8= 748*8= )&38= 1*= 243)*_= '*8= 1*997*8= .3).6:*39= 1*8= ,*37*8_= = a= 5->1147(-.8`= 1=a=1 *5-&1&39-*7&` =!=a= 5.5&(9.8`= '=a= .24)47:2`= 8=a= $*499.&`= =a= &1247(-.8`= .=a ="-&.& _= '*8= 84:8897.':8= (439*3&39= A= 1&= +4.8= )*8= *85T(*8= 2>(4-S9S749745-*8= *3= 7S,.43= 9745.(&1*= *9= )&:97*8=*85T(*8=2>(4-S9S749745-*8=*3=7S,.43=9*25S7S*=8439=+.,:7S*8=*3=47&3,*`=*9=1*8=,*37*8= (477*8543)&398=8439=57S(.8S8=*397*=5&7*39-T8*8_='*8=&:97*8=84:8897.':8=57S8*39*8=)&38=:3=8*:1= )*8=)*:==S(48>89T2*8=8439=+.,:7S*8=*3=34.7_=1*8=).897.':9.438=3*=9.*33*39=(4259*=)*8=*85T(*8= (-14745->11.*33*8=6:*=54:7=1*8=3*499.S*8_=

35 '*8=S9:)*8=57S(S)*39*8=3*=8439=5&8=(425&7&'1*8=*9=3*=5*72*99*39=5&8=)*=7S543)7*= A= (*8= 6:*89.438_= 5:3*= 5&79`= 1*8= (-&25.,3438= 57S(S)*22*39= .)*39.+.S8= (-*?= )*8= 2>(4-S9S749745-*8=9745.(&1*8=439=94:8=S9S=.841S8=A=5&79.7=)*=5*149438=B.,_=/=4:=)*=7&(.3*8= )47(-.)S*8=2>(4-S9S749745-*8=2.8=*3=(:19:7*_=%7=1&=2&/47.9S=)*8=*85T(*8=)*=(-&25.,3438= *(942>(47-.?.*38= 8439= (433:8= 54:7= 3*= 5&8= U97*= (:19.;&'1*8= 5&7= (*8= 2S9-4)*8`= *9= (*8= (:19:7*8=8439=5&7=(4397*=94:9=A=+&.9=5745.(*8=&:=)S;*1455*2*39=)*=8&5745->9*8_=5*=(*=+&.9`= .1=*89=5488.'1*=6:*=1&884(.&9.43=A=)*8=(-&25.,3438=*(942>(47-.?.*38=84.9=7*89S*=(&(-S*=4:= 2&1= )S(7.9*`= /:86:A= (*= 6:*= 1&;T3*2*39= )*8= 2S9-4)*8= )*= '.414,.*= 241S(:1&.7*= 5*72*99*= 1*:7= .)*39.+.(&9.43_= 5&:97*= 5&79`= 1*8= *85T(*8= S9:).S*8= &55&79*3&.*39= A= )*8= 84:8897.':8= ).89.3(9*8`= 6:.= )S7.;*39= 574'&'1*2*39= )S;T3*2*398= ).89.3(98= )&55&7.9.43= )*= 1&= 2>(4-S9S749745-.*= #41;7&>`= ,***_= 4:7= 7*1.*7= 1*8= 24)T1*8= 9*25S7S8= *9= 9745.(&:=`= *9= 7S543)7*=&:==6:*89.438=6:*=1*:78=).++S7*3(*8=4:=1*:78=54.398=(422:38=548*39=.&'1*=/=.1= +&:9=&:=(4397&.7*=8*=51&(*7=)&38=:3=2U2*=(&)7*=5->14,S3S9.6:*_== = = ,_, '*8=3*499.S*8=(422*=24)T1*=)S9:)*= = &72.= 94:9*8= 1*8= *85T(*8= 2>(4-S9S749745-*8= )S(7.9*8`= .*= 8439= )*8= 47(-.)S*8= '*&0*`=+33.`=)&38=1S;41:9.43=)*86:*11*8=1&=2>(4-S9S749745-.*=*89=&55&7:*=A=)*=2:19.51*8= 7*57.8*8=#41;7&>`=,***_='&=51:5&79=)*8=,*37*8=)47(-.)S*8=2>(4-S9S749745-*8=57S8*39*39= 8*:1*2*39=6:*16:*8=*85T(*8`=94:9=&:=51:8=:3*=97*39&.3*='*&0*`=+33.=*9=8439=97T8=).85*78S8= )&38=1&=5->14,S3.*=)*8=47(-.)S*8=B.,_=1_=:=3.;*&:=,S4,7&5-.6:*=&:88.`=(*8=*85T(*8=8439= 97T8= 1&7,*2*39= ).897.':S*8=a= 1&= 2&/47.9S= 8*= 974:;*= *3= 8.*= ):= :)8!89= +,*= *85T(*8`= 349&22*39= &:= <&543`= *9= *3= 2S7.6:*= ):= :)= 2*= *85T(*8_= *:1*8= +/= *85T(*8= 8439= 9*25S7S*8=*9=.1=3*=.89*=6::3=5*:=)*=,*37*8=57S8*398=A=1&=+4.8=*3=2.1.*:=9745.(&1=*9=9*25S7S= 5.54,.:2 `= *5-&1&39-*7& _= !3= )S5.9= )*= (*99*= 7S5&79.9.43`= 1*8= S9:)*8= 439= *88*39.*11*2*39= 5479S=8:7=1*8=47(-.)S*8=2>(4-S9S749745-*8=9*25S7S*8`=(*=6:.=5*:9=(4397.':*7=A=)433*7=:3*= +&:88*= .)S*= ):= +43(9.433*2*39= )*= (*99*= &884(.&9.43_= !11*8= 8439= 94:9*8= &884(.S*8= 97T8= 85S(.+.6:*2*39= A= )*8= (-&25.,3438= *(942>(47-.?.*38= 9*18= 6:*= 1*8= 7:88:1&(S*8`= 1*8= 8*'&(.3&1*8= 4:= 1*8= 9-S1S5-47&(S*8= .&'1*&:= -_= 5&38= 1&= 2*8:7*= 41= 18.*= ):= :)8*89= 7*57S8*39*= :3= 54.39= (-&:)= )*= ).;*78.9S`= 349&22*39= 54:7= 1*8= 47(-.)S*8= 2>(4-S9S749745-*8`= 2&.8= &:88.= 54:7= )&:97*8= 2>(4-S9S749745-*8= .&'1*&:= ,`= /&.= (-4.8.= )S9:).*7=1*8=47(-.)S*8=2>(4-S9S749745-*8=)&38=(*99*=5&79.*=):=243)*_=+.3=)*=54:;4.7= &55479*7= :3*=.39*757S9&9.43=S;41:9.;*= 5*79.3*39*`= /&.=S9:).S= )*8= )*85T(*8= &55&79*3&39=A= )*8=,*37*8=57S8*398=A=1&=+4.8=*3=2.1.*:=9*25S7S=*9=9745.(&1`=9*18=6:*=1*8=,*37*8= *5-&1&39-*7&= *9= 5.54,.:2 =B.,_=1=*9=)*=,*37*8=9745.(&:==574(-*8=)*=,*37*8=9*25S7S8=)S/A=S9:).S8`=9*1= 6:*=1*=,*37*= 5->1147(-.8 =9745.(&1`=574(-*=)*8=,*37*8= $*499.& =*9= *5-&1&39-*7&`= B.,_=1_=

36 5.5&(9.8=97*2418..=Epipactis tremolsii 5.5&(9.8=1:8.9&3.(&=Epipactis lusitanica 5.5&(9.8=1*594(-.1&=Epipactis leptochila 5.5&(9.8=+.'7..=Epipactis fibrii 5.5&(9.8=):7.*38.8=Epipactis duriensis 74 5.5&(9.8=2:*11*7.=Epipactis muelleri 5.5&(9.8=+1&;&=Epipactis flava 5.5&(9.8=-*11*'47.3*=Epipactis helleborine = 5.5&(9.8=2.(745->11&=Epipactis microphylla != 70 5.5&(9.8=&1'.*38.8=Epipactis albiensis 5.5&(9.8=+&,*9.(41&=Epipactis fageticola = 5.5&(9.8=5&1:897.8=Epipactis palustris 100 5->1147(-.8=(&:)&9&=Aphylloçrchis caudata 82 5->1147(-.8=2439&3&=Aphyllorchis montana = .24)47:2=&'479.;:2=Limodorum abortivum = 72 86 *5-&1&39-*7&=143,.+41.&Cephalanthera longifolia =(C) *5-&1&39-*7&=)&2&843.:2=Cephalanthera damasonium 100 *5-&1&39-*7&=*=.,:&=Cephalanthera exigua

*5-&1&39-*7&=7:'7&=Cephalanthera rubra

90 .89*7&=4;&9&=Listera ovata

*499.&=3.):88&;.8=Neottia nidus-avis == 1= 72 .89*7&=82&1..=Listera smallii

"&1247(-.8=97.14':1&9&=Palmorchis '=

94 #-&.&=8&5745->9.(&=Thaia saprophytica =

#745.).&=541>89&(->&=Tropidia polystachya

*7;.1.&=8-.3*38.8=Nervilia shinensis

$&3.11&=51&3.+41.&=Vanilla planifolia

0.03 0.03 substitution / site

= B.,:7*=2_= 7'7*=5->14,S3S9.6:*=)*8=3*499.S*8`=7*(43897:.9=A=5&79.7=)*8=.`= 7'( =*9= 973#8. `=5&7=2&=.2:2= )*=;7&.8*2'1&3(*`=&;*(=+****=7S51.(&98_='*8=*85T(*8=*3=47&3,*=8439=2>(4-S9S749745-*8`=(*11*8=*3='1*:= 8439=2.=49745-*8$=*9=(*11*=*3=34.7=8439=&:949745-*8_= = =

37 '*= ,*37*= 5.54,.:2 = *89= (42548S= )*= (.36= *85T(*8= 2>(4-S9S749745-*8`= )439= 5.54,.:2= 748*:2 `= :3*= 47(-.)S*= &8.&9.6:*= 9745.(&1*_= 1= *89= 51&(S= )&38= 3*7;.1.S*8`= )*8= *5.)*3)74.)S*8=B.,_=1=b=7.),*43= *9=&1 _`=,**/=*9= 5.54,.:2=&5->11:2 `=*:7&8.&9.6:*=9*25S7S*= (S1T'7*=54:7=8&=7&7*9S=)&38=1*=8:)=)*=1!:745*_=5*=2U2*`=1&=548.9.43=5->14,S3S9.6:*=)*8= *85T(*8=) 5.54,.:2 =3&;&.9=5&8=S9S=S9:).S*`=*9=/&.=(43+.72S=6:*=1*8=.3).;.):8=) 5.54,.:2 = &5->11:2 =)*=B7&3(*`=)*=$:88.*=*9=):=<&543=&55&79*3&.*39='.*3=A=1&=2U2*=*85T(*=B.,_=3`= (+_ = &79.(1*= = 54:7= 54:;4.7= (425&7*7= 1*8= (-&25.,3438= 2>(47-.?.*38= )*= (*99*= *85T(*= &:== *=97U2.9S8=)*=843=&.7*=)*=7S5&79.9.43_= '&= 548.9.43= ):= ,*37*= 5->1147(-.8= 3S9&.9= +43)S*= 6:*= 8:7= )*8= )433S*8= 2475-414,.6:*8=*9=S9&.9=97T8=).8(:9S*=57*881*7`=+33*=b=&9*2&3= *9=&1_` =,**/_=<&.=(43+479S=1&= 548.9.43=)*=)*:==*85T(*8=&:=24.38=&:=8*.3=)*8=3*499.S*8=5&7=:3*=5->14,S3.*=241S(:1&.7*= B.,_=2`= &79.(1*= _='*8=,*37*8= *5-&1&39-*7& =*9= 5->1147(-.8 =&55&79.*33*39=A=1&=2U2*=97.':`= 1*8=3*499.S*8=6:.=8439=)*8=*5.)*3)74.)S*8='&8&1*8=B.,_=0`=B.,_=2_=1*99*=97.':=*89=(42548S*= )*3;.743=+.*=*85T(*8=)439=--=8439=2>(4-S9S749745-*8=&9*2&3= *9=&1_` =,**/`=7S5&79.*8= )&38=-=,*37*8=a= $*499.& `= *5-&1&39-*7& =*9= 5->1147(-.8 _=03*=5&79.*=)*8=&:97*8=*85T(*8=;*79*8= 8439=2.=49745-*8$=B.,_=3=*9=(*=97&.9=54:77&.9=U97*=&3(*897&1=(-*?=1*8=3S499.S*8=B.,_=2_= 843= 8*:1*2*39= (*8= ,*37*8= 5*72*99*39= :3*= (425&7&.843= )*8= 2.1.*:== 9*25S7S8= *9= 9745.(&:== )&38= :3= (&)7*= 5->14,S3S9.6:*= 7*897*.39`= 2&.8= 1*8= 97.':8= &:=6:*11*8= .18= &55&79.*33*39=7*57S8*39*39=)*:==-.894.7*8=S;41:9.;*8=.3)S5*3)&39*8_= %1++,-*=_=&5->11:2=7&3(* 2,

++ %1++,-+=_=&5->11:2=7&3(*

+** %1++,,3=_=&5->11:2=0&5&3

2* %1++,,1=_=&5->11:2=*:88.&

2. %1++,-,=_=748*:2

+** -&8974).&=*1&9&

*7;.1.&=8-.7*38.8

+7(-.8=2.1.9&7.8

7&3.(-.8=7*;41:9&

0.8 = B.,:7*= 3_= 48.9.43= 5->14,S3S9.6:*= )*8= ).++S7*398= 5.54,.:2= &5->11:2 = ):= <&543`= $:88.*= *9= B7&3(*= S9:).S8=)&38=1 &79.(1*= _=->14,S3.*=(43897:.9*=5&7= 2&=.2:2=)*= ;7&.88*2'1&3(*=24)T1*= .$=A= 5&79.7= 1*8= .`= &57T8= +***= 7S5S9.9.438_= '*8= 5.54,.:2 = *9= .&8974).& = 8439= 7&88*2'1S8= &1478= 6:*= 1*8= .&8974).& =3&55&79.*33*39=5&8=&:==3*7;.1.S*8=)&57T8=1*8=(1&88.+.(&9.438=9&=4342.6:*8=1-&8*= *9=&1 _`= ,**-_= 1*99*= .3(43,7:*3(*= 3&= 5:= U97*= 7S841:*= 54:7= 1*= 242*39= (&7= &:(:3*= &:97*= 8S6:*3(*= )*= 8*7;.1.S*8= 4:= )*= &8974).S*8= 3*89= ).8543.'1*= 8:7= *3'&30_= 1= 8*7&.9= 3S&324.38= 3S(S88&.7*= )&:,2*39*7= 1S(-&39.11433&,*= 54:7= 7S84:)7*= (*99*= .3(43,7:*3(*_ = '*8= *85T(*8= *3= 47&3,*= 8439= 2>(4-S9S749745-*8`= (*11*= *3= 34.7= 8439= &:949745-*8= *9= 1*8= '7&3(-*8= 748*8= 7*,74:5*39= 1*8= *85T(*8= 2>(4-S9S749745-*8_=

38 = &=

*499.S*8 =;*79*8=2.=49745-*8`=# = 343=.3(1:8*8=2&.8=7'('=*9=5&8= &251.+.&'1*8= 48.9.43=)*8= *7;.1.S*8= )439= 5.54,.:2=&5->11:2 =*9= &8974).S*8 =# = )&57T8=1*8=.=&>&39=5*7):= 7'( =*9= 58&= 343=.3(1:8*8=)&38=1&=5->14,S3.*_=

%7(-.)S*8=;*79*8=574(-*8=)*=# = %7(-.)S*8=2>(4-S9S749745-*8= '= 74:5*=*=9*73*=&:949745-* = 47&1147-.?&=97.+.)& =(4251*= =

47&1147-.?&=2&(:1&9& =(4251*= =

47&1147-.?&=897.&9& =(4251*= =

-&3,*2*398=8:7=1*=,T3*= 7'( = 4)*=)*=;.* = 38*79.43= %7(-.)S*8=;*79*8= 5S1S9.43= %7(-.)S*8=2>(4-S9S749745-*8= 14)43=8945=&39.(.5S= = B.,:7*=+* _=&=->14,S3.*=)*8=%7(-.)S*8=7S&1.8S*=A=5&79.7=):3=,T3*=(-147451&89.6:*`= 58& =*9=548.9.43= )*=6:*16:*8=*85T(*8=2>(4-S9S749745-*8=54:7=1*86:*11*8= 58& =S9&.9=&251.+.&'1*8`=)&57T8=1&2*743`= ,**._= '= ->14,S3.*= ):= ,*37*= 2>(4-S9S749745-*= 47&1147-.?& = 7S&1.8S*= A= 5&79.7= ):3= &:97*= ,T3*= (-147451&89.6:* `= 7'( `= *9= .3).(&9.438= )*8= 57.3(.5&1*8= 24).+.(&9.438= )*= (*= ,T3*`= )&57T8= &77*99= ]= B7*:)*389*.3`=,**2 _=

39 ,_- 5*:==-.894.7*8=S;41:9.;*8=).++S7*39*8=A=(425&7*7= = '-.894.7*=S;41:9.;*=)*=1&=2>(4-S9S749745-.*=(-*?=1*8=47(-.)S*8=*89=2S(433:*=(&7= 1*8= *85T(*8= 2>(4-S9S749745-*8= 3*= 8439= 5&8= 94:/4:78= .3(1:8*8= )&38= )*8= 5->14,S3.*8_= !3= *++*9`=343=8*:1*2*39=(*8=*85T(*8=8439=7&7*8`=2&.8=1*8=,T3*8=:9.1.8S8=*3=5->14,S3.*=3*=8439= 5&8=94:/4:78=&251.+.&'1*8=(&7=.1=8&,.9=)*=,T3*8=(-147451&89.6:*8_== '-.894.7*= S;41:9.;*= )*= 1&= 2>(4-S9S749745-.*= &55&7&9= 51:8= 4:= 24.38= &3(.*33*= 8*143=1*8=97.':8=)47(-.)S*8_='*8= &8974).S*8=*9=8*7;.1.S*8`=9*11*8=6:*= 5.54,.:2 `= .&8974).&`= ,.)>2451*=.8=*9=*(&347(-.8 =3*=8*2'1*39=5&8=&;4.7=:3*=(45.*=):=,T3*= 7'(= 4: =58& =3.=2U2*= )*=(-147451&89*=1&2*743`=,**._=#*8=*=5S7.*3(*8=&:=1&'47&94.7*=)*=1_=)*=&25-.1.8=*33= 9&9*=03.;*78.9>`=0=)&251.+.(&9.43=)*=,T3*8=(-147451&89.6:*8=(-*?= _=&5->11:2 =A=1&.)*= )&247(*8=:3.;*78*11*8=8*=8439=841)S*8=5&7=)*8=S(-*(8=8:7=51:8=)*=,*=,T3*8`=9&3).8=6:.18= S9&.*39=94:8=&251.+.S8=(-*?=:3*=&:97*=2>(4-S9S749745-*`= $*499.&=3.):88&;.8 _=1*99*=5*79*=)*8= (-147451&89*8=4:=):=,S342*=(-147451&89.6:*`=&:=24.38=*3=5&79.*=.3).6:*=6:*=1&=+.=&9.43=)*= 1&= 2>(4-S9S749745-.*= *89= 574'&'1*2*39= :3= S;S3*2*39= &3(.*3`= (422*= (-*?= 1*8= ':72&33.&(S*8= #*70(== *9= &1_` = ,**2_= 5&.11*:78`= (*8= )*:== 84:8897.':8= 8*= 51&(*39= 5->14,S3S9.6:*2*39= 5&72.= 1*8= *5.)*3)74.)S*8= '&8&1*8= )&57T8= 1*8= .`= *9= 57S8*39*39= )*= 143,:*8='7&3(-*8= (+_ &79.(1*= `=1&2*743`=,**.`=(*=6:.=84:1.,3*=:3*=S;41:9.43=143,:*=4:= '.*3=:3*=S;41:9.43=7&5.)*_= '*8=*85T(*8=)*8=,*37*8= >7948.&`=7>9-747(-.8`= 8*:)4;&3.11& `=94:9*8=)*8=;&3.11.4.)S*8= A= 1&= '&8*= )*8= 47(-.)S*8= 5488T)*39= *3= 7*;&3(-*= 7'( = *9= 58& `= *9= 574'&'1*2*39= )*8= (-147451&89*8_= 8S&324.38`= 94:9*8= (*8= *85T(*8= 57S8*39*39= 94:/4:78= )*= 143,:*8= '7&3(-*8= 1&2*743`= ,**.=b= B.,_= +*= *9= (*8= ,T3*8= (-147451&89.6:*8= 439= 574'&'1*2*39= 8:'.= )*= 342'7*:8*8=2:9&9.438=343=(4397*88S1*(9.433S*8`=6:.=*3=439=+&.9=)*8=58*:)4,T3*8_== !3+.3`=1*8=*85T(*8=):=,*37*= 47&1147-.?& =5488T)*39=94:/4:78=:3*=(45.*=)*= 7'( `= 58& = *9= 2&96= 2&.8= 1*:7= ,T3*= 7'( = 5488T)*= '*&:(4:5= 24.38= )*= 24).+.(&9.438= 6:*= (*1:.= )*= >7948.& =85_=&77*99=]=B7*:)*389*.3`=,**2_='*8= 47&1147-.?& =57S8*39*39=94:/4:78=)*=143,:*8= '7&3(-*8= )&38= 1*8= 5->14,S3.*8= B.,_= +*_= 1*5*3)&39`= 47&1147-.?&= 97.+.)& `= :3*= 2>(4-S9S749745-*=1S,T7*2*39=;*79*=*89=(&5&'1*=)*=5-4948>39-T8*=Z.22*7= *9=&1_` =,**2`=*9= 843=,T3*= 7'( =3*=57S8*39*=5&8=)*=24).+.(&9.43=2&/*:7*=B.,_=+*_= 5&38= 1&= 97.':= )*8= 3*499.S*8`= $*499.&= 3.):88&;.8 = 5488T)*=&:88.= :3= (*79&.3=3.;*&:= )*= (-14745->11*=+*.+-4+*7`=+323`=&.38.=6:*= .24)47:2=&'479.;:2= .71&3)&= *9=&1_` =,**0_=5&38= (*99*=2U2*=97.':`=94:9*8=1*8=*85T(*8=2>(4-S9S749745-*8=):=,*37*= *5-&1&39-*7& =57S8*39*39= )*8=+*:.11*8=51:8=)S;*1455S*8=6:*=(-*?=94:9*8=1*8=&:97*8=2>(4-S9S749745-*_='&=5->14,S3.*= 241S(:1&.7*= )*= (*99*= 97.':= 3*= 57S8*39*= 5&8= )*= 143,:*8= '7&3(-*8= (&7&(9S7.89.6:*8= )*8= 2>(4-S9S749745-*8_= '&(6:.8.9.43= )*= 1&= 2>(4-S9S749745-.*= (-*?= 1*8= 3*499.S*8= 7S8:19*7&.9= )43(= ):3=S;T3*2*39= 51:8= 7S(*39`= 4:= '.*3= 1S;41:9.43= )*=(*99*= 97.':= 8*7&.9= 24.38=7&5.)*= 6:*=1*8=&:97*8_== &38= 54:;4.7= (43(1:7*= 8:7= 1&3(.*33*9S= )*= 1&(6:.8.9.43= )*= 1&= 2>(4-S9S749745-.*= )&38= (*8= 84:8897.':8`= )*8= ).++S7*3(*8= )*= 7>9-2*= )S;41:9.43= 8439= S;.)*39*8= *9= /:89.+.*39= )&:9&39=51:8=1&=3S(S88.9S=)*=8*=51&(*7=)&38=:3=2U2*=(&)7*=5->14,S3S9.6:*=54:7=(425&7*7= 1*8=24)T1*8=9*25S7S8=*9=9745.(&:==

40 & *7;.1.*&* = *7;.1.&= _=&5->11:2 5.54,.:2= _=748*:2= 457.3&(S*8 5.1;47(-.8= 59*7*48&3)7&= 884(.&9.43= &974).*&* = 85S(.+.6:*=A=)*8= -&8974).&= 72.11&7.&= 8&5745->9*8=4:= %1*.47(-.8= 42*8= )*8=5&7&8.9*8 *4(1*2*38.&= 3*749:8 .)>2451*=.*11&= 2&7&82.:8 .)>2451*=.8= :=45:8= =9745.(&1*= =9*25S7S= ' :949745-*=9745.(&1*

.9*=)S9:)*=) _=748*:2 == .9*=)S(-&39.11433&,*== ) _=&5->11:2

(

884(.&9.43=85S(.+.6:*=)*== 5.54,.:2=&5->11:2= A=)*8=(-&25.,3438= = ):=,*37* =434(>'*= 85_`== 2&/47.9&.7*2*39== 434(>'*=+:8(.):1&

= B.,:7*= ++_= 48.9.43= .&=4342.6:*= ) 5.54,.:2= &5->11:2 = *9= 2>(47-.?*8= )*= 8*8= 574(-*8= 5&7*398= 2>(4-S9S749745-*8= &`= 1&= 5->14,S3.*= 3*89= 5&8= 7S841:*= *397*= 94:9*8= (*8= *85T(*8`= (1&88.+.(&9.43= )&57T8=1-&8*= *9=&1_` =,**-_=7.3(.5&:==7S8:19&98=.88:8=)*=1 &79.(1*= ='=*9=.11:897&9.43=)*= _=&5->11:2 = (_= '*8= 1*997.3*8= 8:7= 5.54,.:2= &5->11:2 = .3).6:*39=a= (7`= 7-.?42*= (47&114)*b= .`= .3+147*8(*3(*b= .'`= '4:943=)*=.3+147*8(*3(*=b=.8`=9.,*=5479&39=1.3+147*8(*3(*b=8`=894143=+.3b=8'`=':1'.11*=8:7=:3=894143_=

41 ,_. :=(4397&.7*=)*8=3*7;.1.S*8=9745.(&1*8`= 5.54,.:2=&5->11:2 = *89=&884(.S*=85S(+.6:*2*39=A=)*8=(-&25.,3438= *(942>(47-.?.*38= = 03*= 57*2.T7*= +&O43= )*= (425&7*7= 1*8= &884(.&9.438= 2>(47-.?.*33*8= )*8= 2>(4-S9S749745-*8=*397*=1*=2.1.*:=9*25S7S=*9=1*=2.1.*:=9745.(&1=*89=)*=8*=5*3(-*7=8:7=)*8= *85T(*8=9*25S7S*8=)439=)*8=574(-*8=5&7*398=9745.(&:==439=)S/A=S9S=S9:).S8_=&7=*=*251*`= 8&(-&39= 6:*= 5.54,.:2= 748*:2 = *89= &884(.S= 85S(.+.6:*2*39= A= )*8= 1457.3&(S*8`= )*8= 8&5745->9*8= B.,_= ++`= &&2&94= *9= &1_` = ,**/=b= &&,&2*= *9= &1_` = ,**2&`= 1*8= 6:*89.43= )&38= 3497*= (439*=9*= )S9:)*= 8439= )*= )S9*72.3*7= 8.= _= &5->11:2 = *89= &884(.S= &:88.= .= A= )&:97*8= 8&5745->9*8=*9=..=8.=(*99*=&884(.&9.43=*89=85S(.+.6:*_=' &79.(1*= =7S543)=A=(*8=6:*89.438=*9= 24397*=6:*=94:8=1*8=.3).;.):8=) _=&5->11:2 =S9:).S8`=&:=<&543`=*3=$:88.*=*9=*3=B7&3(*`=8439= &884(.S8= A= )*8= *85T(*8= ):= ,*37*= !34(>'* `= :3= ,*37*= *(942>(47-.?.*3`= 8&38= 85S(.&1.8&9.43= ,S4,7&5-.6:*=B.,_=++_=5&:97*8=(-&25.,3438=51:8=7&7*8=439=S9S=.)*39.+.S8=*9=94:8=S9&.*39= *(942>(47-.?.*38_= 1= 8*2'1*= )43(= 6:*= 1&= ,S3S7&1.8&9.43= ):= 24)T1*= 85S(.+.(.9S= (= *(942>(47-.?.*3= 84.9= 89&'1*= *3= 2.1.*:= 9*25S7S`= *9= 6:.1= 3>= &.9= 5&8= )*= (4397&.39*= 5->14,S3S9.6:*= 6:.= )S9*72.3*= 1.)*39.9S= *9= 1S(414,.*= )*8= (-&25.,3438= &884(.S8= (-*?= 1*8= 5.54,.:2 _= '*8= )*:== *85T(*8= ) 5.54,.:2 = 8439= &884(.S*8= 85S(.+.6:*2*39= A= )*8= (-&25.,3438=).++S7*398=343=8*:1*2*39=)*=5&7=1*:7=S(414,.*=2&.8=&:88.=)*=5&7=1*:7=548.9.43= 5->14,S3S9.6:*_=5*=51:8`=1*8= _=&5->11:2 =):=<&543=8439=*:==&:88.=&884(.S8=A=)*8=.34(>'*8`= (*=6:.=24397*=6:*=)*8=47(-.)S*8=&8.&9.6:*8`=4:=94:9=):=24.38=(*11*8=54:88&39=*3=2.1.*:= 9*25S7S=5*:;*39=U97*=&884(.S*8=A=)*8=(-&25.,3438=*(942>(47-.?.*38_== = = ,_/ 1422*= 1*8= 3*499.S*8= 9*25S7S*8`= 1*8= 3*499.S*8= 9745.(&1*8= 8439=&884(.S*8=A=)*8=(-&25.,3438=*(942>(47-.?.*38= = '&=6:*89.43=.3;*78*=*89=)S9:).*7=*3=2.1.*:=9745.(&1=)*8=*85T(*8=2>(4-S9S749745-*8= )439= )*8= 574(-*8= 5&7*398= 439= S9S= S9:).S8= *3= 2.1.*:= 9*25S7S_= &7= *=*251*`= (-*?= 1*8= 3*499S*8`= *5-&1&39-*7&= &:89.3&* = *89= &884(.S*= 85S(.+.6:*2*39= A= )*8= .-S1S5-47&(S*8= *(942>(47-.?.*33*8= &:== !9&98803.8= .&>147= ]= 7:38`= +331`= $*499.&= 3.):88&;.8 = A= )*8= 8*'&(.3&1*8=*(942>(47-.?.*33*8=*3=!:745*=*1488*= *9=&1_` =,**,'_=I:*11*=*89=1&=8.9:&9.43=*3= 8.*=9745.(&1*`=7*85*(9.;*2*39`=54:7=1*85T(*= *5-&1&39-*7&=*=.,:& =*9=1*=,*37*= 5->1147(-.8= = ' &79.(1*= =7S543)=A=(*99*=6:*89.43=*3=.-&1&3)*`=*9=24397*=6:*= _=*=.,:&`=_=2439&3& =*9= _= (&:)&9& =8439=94:9*8=974.8=&884(.S*8=A=)*8=(-&25.,3438=*(942>(47-.?.*38=B.,_=+,_=5&:97*8= &79.(1*8= 8439= ;*3:8= (43+479*7= (*8= )S(4:;*79*8= .&'1*&:= -=a= -&2&*,&8974).&= 8.040.&3& `= :3*= 47(-.)S*=2>(4-S9S749745-*=S9:).S*=*3=147S*=2&.8=57S8*39*=)&38=94:9*=18.*=):=:)8*89`= *89=5&7=*=*251*=&884(.S*=A=)*8= *7&94'&8.).:2 =):3=(1&)*=*(942>(47-.?.*38=&&,&2*= *9=&1_` = ,**2'_= 5&.11*:78`= /&.= &:88.= .)*39.+.S= )*8= *7&94'&8.).:2= )433S*= 343= 5:'1.S*= )&38= 1*8= 7&(.3*8= ):3= 85S(.2*3= )*= -&2&*,&8974).&= 54.1&3*. = 7S(419S= 57T8= )*= 54.= 39-&343`= *3= .-&1&3)*`= )&38= )*8= +47U98= -:2.)*8= )&19.9:)*= A= B&,&(S*8`= 2&.8= 1*= +&.'1*= 342'7*= )S(-&39.11438=3*=5*72*9=5&8=)*=,S3S7&1.8*7=54:7=(*99*=*85T(*_== =

42 &= = =)*8=(-&25.,3438=2>(47-.?.*38=.)*39.+.S8 = 100% (414,.*=(433:*=)*8=(-&25.,3438= .)*39.+.S8=)&38=1*8=7&(.3*8=)*8=974.8= 80% *85T(*8=2>(4-S9S749745-*8=S9:).S*8=

1-&25.,343=!1#= 60% $-.?4(943.&= !3)45->9*= 40% &5745->9*=

20%

0% AM AC CE 10 '= 8 _=(&:)&9& = +/ 8== 6 *3== 4 :).8.&=).8(4147 = 2

-2

-4 (= 14 12 +/ 8== 10 _=*=.,:& = *3== 8 6 4 2 0 -*.7489>1.8=2439&3& = -2 -4 -6 -40 -35 -30 -25 -20

+- 1=*3= = 9&9:9=9745-.6:*=)*8=S(-&39.11438 = == = 1-&25.,343=*(942>(47-.?.*3= = %7(-.)S*=2>(4-S9S749745-*= = %7(-.)S*=9*77*897*=343=2>(4-S9S749745-*= = :949745-*8=8&:+=47(-.)S*8= = B.,:7*= +,_= &= !(414,.*= )*8= (-&25.,3438= 2>(47-.?.*38= )*= 5->1147(-.8= 2439&3& = #`= _= (&:)&9& = 1=*9= *5-&1&39-*7&=*=.,:& =1!_='`=(=.,3&9:7*=.84945.6:*=)*= _=(&:)&9& =*9= _=*=.,:& =*9=)*=)*:== 47(-.)S*8= 9*77*897*8`= :).8.&= ).8(4147 = *9= -*.7489>1.8= 2439&3& = (425&7S*8= A= (*11*8= )*8= 51&39*8= &:949745-*8=*9=)*=(-&25.,3438=*(942>(47-.?.*38_=5&57T8=1 &79.(1*= _=

43 5&38= (*8= )*73.*78= (&8`= 1*8= (-&25.,3438= *(942>(47-.?.*38= .)*39.+.S8= 8439= )*8= *7&94'&8.).:2 = 4:= )*8= #*'&(.3& =a= )*8= ,74:5*8= 7&88*2'1S8= 5&7= &.11*:78= 84:8= 1*= 9*72*= )*= 7-.?4(943.&$`= )-&'.9:)*= (438.)S7S8= (422*= )*8= ,74:5*8= )*= 8&5745->9*8= (&5&'1*8= )*= 8&884(.*7=8>2'.49.6:*2*39=:3.6:*2*39=&:==47(-.)S*8_=1*8=)S(4:;*79*8=24397*39=6:*=.= 1*8= 7-.?4(943.&8$= 439= :3*= S(414,.*= 5&7+4.8= *(942>(47-.?.*33*`= ..= .18= 8439= 97T8= 7S5&3):8= (422*=1*8=*'&(.3&1*8=b= 1*3= *9=&1_ `=,**,=b=*1488*= *9=&1 _`=,**,'=b=F*.88= *9=&1 _`=,**.`=*9=...=1*= (-&3,*2*39= )*= (-&25.,3438= 3*89= 5&7+4.8= 5&8= 8.= ,7&3)= ):= 54.39= )*= ;:*= 9&=4342.6:*= *397*= )*8= 47(-.)S*8=;*79*8=*9= 1*8= 2>(4-S9S749745-*8=(43(*73S*8_= &7= (4397*`= ):= 54.39= )*= ;:*=S(414,.6:*`=.1=8&,.9=):3=(-&3,*2*39=2&/*:7_=1*99*=574=.2.9S=9&=4342.6:*=)*85T(*8= )*= (-&25.,3438= 7-.?4(943.&$= *9= )*= (-&25.,3438= *(942>(47-.?.*38= 54:77&.9= +&(.1.9*7= 1&= 97&38.9.43= ;*78= 1&= 2>(4-S9S749745-.*`= :3= S;S3*2*39= +7S6:*39= (-*?= 1*8= 47(-.)S*8`= 349&22*39=9745.(&1*8_= = ,_0 B1:==)*=(&7'43*=*9=1#8=84:8=1*8=9745.6:*8== '*8= 3*499.S*8= 2>(4-S9S749745-*8= &8.&9.6:*8= 8439= &884(.S*8= A= )*8= (-&25.,3438= *(942>(47-.?.*38`=*9=*3=7*O4.;*39=94:9=1*:7=(&7'43*=47,&3.6:*`=(422*=84:1.,3*=1&3&1>8*= )*= 1*:7= 8.,3&9:7*= .84945.6:*= *3= +- 1= *9= +/ 8= B.,_= +,'`= (+= &79.(1*= `= 574(-*= )*= (*11*= )*8= (-&25.,3438=*(942>(47-.?.*38_=.38.`=(*8= 47(-.)S*8=2>(4-S9S749745-*8= (4389.9:*39= :3*= 57*:;*= .3).7*(9*= )*8= +1:== )*= (&7'43*= *397*= 1*8= 51&39*8= *9= 1*8= (-&25.,3438= *(942>(47-.?.*38_= '*8= 1#8= 439= 94:/4:78= S9S= S9:).S8= *3= 2.1.*:= 9*25S7S`= *9= 8.18= 8439= 549*39.*11*2*39=57S8*398=)&38=)*8=+47U98=A=5.59*74(&75&(S*8=*9=B&,&(S*8`=1*:7=*=.89*3(*=3>= &=/&2&.8=S9S=)S24397S*_='&=57S8*3(*=)*=(*8=47(-.)S*8=(4389.9:*=)43(=:3=57*2.*7=&7,:2*39= *3=1*:7=+&;*:7_=5&.11*:78`=(*8=47(-.)S*8=2>(4-S9S749745-*8=(48*=.89*39=84:;*39=&;*(=)*8= 47(-.)S*8= 9*77*897*8=343= 2>(4-S9S749745-*8_= !3= .-&1&3)*`= ,*= )*8=*85T(*8= )47(-.)S*8= 8439=9*77*897*8`=*9=54:7=1&=51:5&79=+47*89.T7*8=*.)*3+&)*3`=+312_=!11*8=54:88*39=)43(=)&38= :3=*3;.7433*2*39=41=5&7+4.8=8*:1=+=)*=1&=1:2.T7*=5&7;.*39=&:=841=.:73':11= *9=&1 _`=+33-`= (*=6:.=84:1T;*=1&=6:*89.43=)*=1*:7=89&9:9=9745-.6:*=*9=)*=1*:7=1.*3=&;*(=1*:78=(-&25.,3438= 2>(47-.?.*38_= '&=8.,3&9:7*=.84945.6:*=B.,_=.=)*8=+*:.11*8=)*=(*8=47(-.)S*8=(-14745->11.*33*8=5*:9= .3).6:*7= 1&= 84:7(*= )*= (&7'43*= 47,&3.6:*= 6:*11*8= :9.1.8*39_= 1= 5*:9= 8&,.9= *39.T7*2*39= )*= (&7'43*= .88:= )*= 1&= 5-4948>39-T8*`= 4:= )*= (&7'43*= 5&79.*11*2*39= .88:= )*8= (-&25.,3438`= A= 1.389&7=)*=(*79&.3*8=3*499.S*8=;*79*8=2.=49745-*8$=*1488*= *9=&1_` =,**0_='&=(425&7&.843=)*= 1*:7= 8.,3&9:7*= .84945.6:*= *3= +- 1= A= (*11*= )*= 51&39*8= &:949745-*8= *9= A= (*11*= )*8= 51&39*8= (4251T9*2*39= 2>(4-S9S749745-*8= 5*72*9= )*= (&1(:1*7= :3= 54:7(*39&,*= )*= (&7'43*= +43,.6:*=)&38=1*8=+*:.11*8`=&55*1S=54:7(*39&,*=)-S9S749745-.*=)&57T8=:3=24)T1*=1.3&.7*= )*=2S1&3,*=)*=)*:==84:7(*8`=-.11.58=]= 7*,,`=,**+ _=1=*89=51:8=)S1.(&9=):9.1.8*7=1*= +/ 8= (422*= 7S+S7*3(*`= 6:.= 57S8*39*= 5&7+4.8= :3*= 8.,3&9:7*= 57457*= A= 1&= +&2.11*= 5&<843= *9= &1_` = ,**,_= 848= 7S8:19&98= 24397*39= 6:*= 1*8= )*:== 47(-.)S*8= 57S8*39*8= 8:7= 1*8= 8.9*8= )S(-&39.11433&,*= )*= _= (&:)&9& = *9= _= *=.,:& = 439= )*8= 8.,3&9:7*8= .84945.6:*8= ).++S7*39*8= 1:3*= )*= 1&:97*=a= :).8.&= ).8(4147 = &= :3*= 8.,3&9:7*= 343= 8.,3.+.(&9.;*2*39= ).++S7*39*= )*8= &:949745-*8= 9&3).8= 6:*= -*.7489>1.8= 2439&3& = &= :3*= 8.,3&9:7*= 51:8= 574(-*= )*= (*11*= )*8= (-&25.,3438_= .38.= (*99*= )*73.T7*= 7*(*;7&.9= 2*= )*= 843= (&7'43*= 47,&3.6:*= )*8= (-&25.,3438=B.,_+,'`= (+ =&79.(1*= _=!11*=8*7&.9=)43(=5&79.*11*2*39=)S5*3)&39*=)*=(*:=8(.`= )43(=2.=49745-*$=8=2&.8=8*8=5&79*3&.7*8=2>(47-.?.*38=3439=5&8=*3(47*=S9S=.)*39.+.S8_= =

44 &_= &=3:97.9.43=)*8=51&39*8=2>(4-S9S749745-*8 =

7(-.)S*= :949745-* = 2>(4-S9S749745-*= -&25.,343= =

'_= &=3:97.9.43=)*8=51&39*8=2.=49745-*8 = 7(-.)S*= 2.=49745-* = :949745-* = -&25.,343= =

:2.T7*=(= ,bbb=5-4948>39-T8* = B1:==)*==47,&3.6:*= B1:==)*&:=*9=)*=8*18=2.3S7&:==`= = = = = = B.,:7*=+-_= (-S2&8=(425&7&9.+=):=+43(9.433*2*39=)*=1&=2>(4-S9S749745-.*=&=*9=)*=1&=2.=49745-.*= '_= '&= 2.=49745-.*= *89= )S+.3.*= 5&7= 1:9.1.8&9.43= )*= )*:== 84:7(*8= )*= (&7'43*= ).++S7*39*8`= .(.= .347,&3.6:* =;.&= 1&=5-4948>39-T8*=*9=47,&3.6:* =;.&= 1*8=(-&25.,3438_ ==

45 '*8= 7&(.3*8= )&:97*8= 47(-.)S*8= (-14745->11.*33*8= 9-&1&3)&.8*8= 439= 7S;S1S= 1&= 57S8*3(*= )&:97*8= (-&25.,3438= *(942>(47-.?.*38`= 349&22*39= )*8= 7:88:1&(S*8= (-*?= "745.).&= (:(:1.3,4.)*8 = *9= *5-&1&39-*7&= 143,.+41.&= (+_ &79.(1*= `= )*8= *1.42>(*8 = (-*?= _= 143,.+41.& = *9= )*8= 5*?.?&1*8= *(942>(47-.?.*33*8= (-*?= #.7.3)-473.&= 5:1(-*11&= )433S*8= 343= 5:'1.S*8_= 5*= 2U2*`= 9*7489>1.8=3:9&38 =*89=&884(.S=85S(.+.6:*2*39=A=)*8= *7&94'&8.).:2 =*(942>(47-.?.*38= *3=:897&1.*=7<.3= *9=&1_` =,**1_== 1*8= 57*2.*78= S1S2*398= 8:,,T7*39= 6:*= 1&= 2.=49745-.*= 5:.88*= U97*= 574'&'1*2*39= 7S5&3):*= )&38= 1*8= +47U98= 9745.(&1*8`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`=34:8=8:,,S7438= 6:*16:*8= (7.9T7*8= ).3;*89.,&9.438= 54:7= .)*39.+.*7= )&:97*8= 2.=49745-*8_= 1= 8*7&.9= 5&7= *=*251*=/:).(.*:==)*=&2.3*7=a= 8 )*8=*85T(*8=5->14,S3S9.6:*2*39=574(-*8=)*85T(*8=2>(4-S9S749745-*8= 8 )*8=*85T(*8=&884(.S*8=A=)*8=(-&25.,3438=*(942>(47-.?.*38= 8 )*8=*85T(*8=)*=84:88'4.8=5&79.(:1.T7*2*39=842'7*= 8 )*8=*85T(*8=&5->11*8=4:=A=+*:.11*8=7S):.9*8= 8 )*8=*85T(*8=)439=1*8=,7&.3*8=8439=)*=5*9.9*8=9&.11*8=*9=439=5*:=)*=7S8*7;*8= 8 )*8=*85T(*8=)439=1*=,&2S945->9*=*89=343=(-14745->11.*3_= = 'S9:)*=)*=1&=2.=49745-.*=8*89=(43(*397S*=(-*?=1*8=47(-.)S*8=*9=1*8=*7.(&(S*8`=9&3).8=6:*= 1*8=2>(4-S9S749745-*8=&55&79.*33*39=&:88.=A=)&:97*8=+&2.11*8=)&3,.485*72*8=4:=2U2*= )*= '7>45->9*8= *9= )-S5&9.6:*8_= 5*= 51:8= &251*8= 7*(-*7(-*8= 8*7&.*39= A= 1&3(*7= )&38= (*= )42&.3*`=*9=5*72*997&.*39=)&2S1.47*7=1*8=2*8:7*8=)*=(438*7;&9.43=)*=9*11*8=*85T(*8`=)&38= 1&=2*8:7*=41=1*8=&884(.&9.438=2>(47-.?.*33*8=8439=7&7*2*39=57.8*8=*3=(4259*_= = *8= 57*2.*78= 7S8:19&98= 34:8= 439= 5*72.8= )*= 7S543)7*= A= (*79&.3*8= 6:*89.438= 84:1*;S*8=5&7=1&=(425&7&.843=)*8=24)T1*8=9*25S7S8=*9=9745.(&:==&'1*=0a== +_ &884(.&9.43= A= )*8= (-&25.,3438= *(942>(47-.?.*38= *89= (43+.72S*= *3= 2.1.*:= 9*25S7S= (422*= *3= 2.1.*:= 9745.(&1_= *8= (4397&.39*8= 5->14,S3S9.6:*8= 8:7= 1.)*39.9S=*9=1S(414,.*=)*8=5&79*3&.7*8=;&7.*=8*143=1*8=97.':8_= ,_ 1&=85S(.+.(.9S=*89=(43+.72S*=*3=2.1.*:=9*25S7S= -_ 1&=2.=49745-.*=*89=5488.'1*=*3=2.1.*:=9745.(&1_= = *5*3)&39`=)*8=6:*89.438=34:;*11*8=439=S2*7,S=)*=1S9:)*=)*8= 5->1147(-.8 _= = 7(-.)S*8= 884(.&9.43=A== .1.*:= .1.*:= )*8=(-&25.,3438= *25S7S= 9745.(&1= >(4-S9S749745-*8= (942>(47-.?.*38= (((= (((= &5745->9*8= 8= (((= &7&8.9*8= 8= (((= 5S(.+.6:*8= (((= (((= .=49745-*8= 5S(.+.6:*8= (=4:=8= 8= (942>(47-.?.*38= (((= (= Tableau 6. Comparaison des modèles tempérés et tropicaux d’association des orchidées mycohétérotrophes et mixotrophes

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

47 _ '&=343885S(.+.(.9S=)*8= 5->1147(-.8 = = = -_+ 03*=8:757.8*=)&38=1*8=7&(.3*8=)*8= 5->1147(-.8 = = '*8= *85T(*8= _= 2439&3& = *9= _= (&:)&9& = 8439= (*79*8= &884(.S*8= A= )*8= (-&25.,3438= *(942>(47-.?.*38`= 2&.8=A= 1&= ).++S7*3(*= )*= 94:9*8= 1*8=&:97*8= 2>(4-S9S749745-*8`=(*8= )*:== 5->1147(-.8 =8439=&884(.S*8=A=1&=+4.8=A=)*8=7:88:1&(S*8`=)*8=9-S1S5-47&(S*8`=)*8=8*'&(.3&1*8`= )*8=(1&;:1.3&(S*8`=)*8=(479.3&7.&(S*8`=)*8=&2&3.9&(S*8_=.38.`=:3=2U2*=.3).;.):=5*:9=U97*= &884(.S= A= 51:8.*:78= *85T(*8= 343= &55&7*39S*8= )*= (-&25.,3438= *3= 2U2*= 9*258`= *9= 1*8= .3).;.):8=;4.8.38=3*=5&79&,*39=5&8=3S(*88&.7*2*39=)*=(-&25.,3438_=1=8&,.9=)43(=):3=(&8= 41= .1= 3*=.89*= 3.= 85S(.&1.8&9.43= 14(&1*`= 3.= 5&79*3&.7*= 2>(47-.?.*3= 4'1.,&94.7*= )&38= (*99*= &884(.&9.43`= *9= )43(= &:(:3*= 85S(.+.(.9S= B.,_= +.`= *9= &79.(1*= _= '&= (422:3&:9S= )*= (-&25.,3438= .)*39.+.S8= )&38= 1*8= 7&(.3*8= )*8= 5->1147(-.8 = *89= )42.3S*= 5&7= )*8= 9&=438= *(942>(47-.?.*38`=6:.=3*=5*:;*39=U97*=(438.)S7S8=(422*=)*8=(439&2.3&398=)*=1&'47&94.7*= 4:= .88:8= ):= 841`= &:= (4397&.7*= )*8= 8&5745->9*8_= 5*= 51:8`= 1*8= 7&(.3*8= 439= S9S= 1&;S*8= 84.,3*:8*2*39=&;&39=)U97*=8S(-S*8_=!3+.3`=1&251.+.(&9.43=)*8=.=A=5&79.7=)*=5*149438=&= S9S= 7S&1.8S*= 8:7= :3= 5*9.9= 342'7*= )S(-&39.11438`= *9= 51:8.*:78= (-&25.,3438= *(942>(47-.?.*38= ).++S7*398= 439= S9S= .)*39.+.S8= (+_ &79.(1*= = (*= 6:.= ;&1.)*= 1*8= 7S8:19&98= *9= 7S+:9*=1->549-T8*=):3*=(439&2.3&9.43=2*3&39=A=(*99*=&'8*3(*=)*=85S(.+.(.9S`=(422*=(*= +29=1*=(&8=54:7= #&7(4)*8=8&3,:.3*& =1:11.3,8= *9=&1_` =+330=b=A7*9?*7= *9=&1_` =,***_=5&:97*=5&79`= (*79&.38=.3).;.):8=3*=5&79&,*39=&:(:3*=*85T(*=)*=(-&25.,343=*3=(422:3`=*9='.*3=6::3*= S9:)*=):=8>89T2*=7&(.3&.7*=(4251*9=84.9=3S(*88&.7*=54:7=(43(1:7*`=43=5*:9=*3=)S):.7*=6:.1= 3>=&=5&8=)*=85S(.+.(.9S=+43(9.433*11*=897.(9*=;.88A8;.8=):3*=*85T(*_=!3+.3`=1&=(425&7&.843= )*8= 8S6:*3(*8= )*= $:88:1&(S*8= &251.+.S*8= A= 5&79.7= )*8= 7&(.3*8= 24397*= 6:.1= 3>= &= 5&8= )*= 85S(.&1.8&9.43=a=1*8=8S6:*3(*8=8*=).85*78*39=)&38=1&7'7*=5->14,S3S9.6:*=)*8=$:88:1*8`=8&38= 7*,74:5*2*39=,S4,7&5-.6:*=3.=2U2*=9&=4342.6:*`=5:.86:*=1*8=8S6:*3(*8=4'9*3:*8=54:7= _=2439&3& =*9= _=(&:)&9& =8439=2S1&3,S*8= (+_ &79.(1*= _== = = -_, '&=343885S(.+.(.9S=a=:3*=5&79.(:1&7.9S=9745.(&1*== = 1=*89='.*3=(433:=6:*=1*=342'7*=)*85T(*8=*89=51:8=,7&3)=84:8=1*8=9745.6:*8`=54:7= ).;*78= ,74:5*8= )47,&3.82*8= *9= 349&22*39= 54:7= 1*8= 51&39*8`= 54:7= 1*8= (-&25.,3438= *3)45->9*8=7341)= *9=&1 _`=,***=b=7341)=]=':9?43.`=,**1=2&.8=5&8=3S(*88&.7*2*39=54:7= 1*8= (-&25.,3438= *(942>(47-.?.*38= 5&-1'*7,`= ,**+=b= 8&7&`= (422:3.(&9.43= 5*78433*11*_= 5*=51:8`=1&=).897.':9.43=):=89&9:9=)*8=*85T(*8`=,S3S7&1.89*=4:=85S(.&1.89*`=6:*=(*=84.9=):3*= 84:7(*= )*= (&7'43*`= ):3= -&'.9&9`= *89= 2S(433:_= 5*= 7S(*39*8= S9:)*8= 8:7= )*8= .38*(9*8= 5->945-&,*8=439=&:88.=84:1.,3S=6:.1=>=&;&.9=51:8=)*=,S3S7&1.89*8=)&38=)*8=+47U98=-:2.)*8= 9745.(&1*8= 6:*3= 2.1.*:= 9*25S7S= 84;493>= *9= &1_` = ,**,= *9= 6:*= 1*= 342'7*= 51:8= .25479&39= )*85T(*=S9&.9=(477S1S=A=1&=).;*78.9S=)*8=*85T(*8=-9*8=57S8*39*8`=2&.8=)*8=(4397*8*=*251*8= *=.89*39=&:88.=(-*?=1*8=.38*(9*8=84;493>= *9=&1_` =,**0=*9=&:(:3*=9*3)&3(*=,S3S7&1*=3*=5*:9= U97*=84:1.,3S*_=

48 & 842'7*=)*85T(*8= 1* 0* /* .* -* ,* +* * 9-*1.&(S*8 *149.&(S*8 *'&(.3&1*8 $:88:1&(S*8 &:97*8=!1# 2&3.9&(S*8 11&;:1.3&(S*8 1479.3&7.&(S*8 14197.(.*11&(S*8 .-*1*5-47&(S*8 B&2.11*8=)*=(-&25.,3438=

' ( 842'7*=)*85T(*8= 40 842'7*=)*85T(*8= +. 35 +, 30 +* 25 20 2

15 0 10 .

5 , 0 * :.11&(S*8 41*9&(S*8 41*9&(S*8 $:88:1&(S*8 $:88:1&(S*8 $&2&7.&(S*8 2&3.9&(S*8 11&;:1.3&(S*8 1479.3&7.&(S*8 1479.3&7.&(S*8 .-*1*5-47&(S*8 1&39-&7*11&(S*8 !394142&9&(S*8 (1*74)*72&9&(S*8 (1*74)*72&9&(S*8 >2*34,&89*7&(S*

B&2.11*8=)*=(-&25.,3438= = = = = = B.,:7*= +/_= 842'7*= )*= 9&=438= )*= (-&25.,3438= *(942>(47-.?.*38= 7&88*2'1S*8= 5&7= +&2.11*= *9= 7*(*38S8=)&57T8=1*:7=.)*39.+.(&9.43=)&38=1*8=7&(.3*8=) _=2439&3& =&`=1*:78=(&7545-47*8='`=)&57T8= 5*11`= ,**/`= 1*:78= *(942>(47-.?*8= (`= )&57T8= &<&824735.9&0= *9= &1_` = ,**0= )&38= )*8= +47U98= )*= 5.59*74(&75&(S*8=):=847)=)*=1&=.-&1&3)*=57T8=)*=1-.&3,=#&._=1-&6:*=+&2.11*=*89=7*57S8*39S*=5&7= :3=+.,:7S=).++S7*39_=

49 8S&324.38`=*3=(*=6:.=(43(*73*=1*8=47(-.)S*8=2>(4-S9S749745-*8`=(*8=)*:==(&8=)*= 343=85S(.+.(.9S=(4397*).8*39=1&=;.8.43=(422:3S2*39=&)2.8*=(43(*73&39=1&=85S(.+.(.9S=)*8= 2>(4-S9S749745-*8= *=51.6:S*= 5&7= 1&= (48&)&59&9.43= +43(9.433*11*= 4:= 1&= (4S;41:9.43= 5&7&8.9&.7*= (+= 5&7&,7&5-*= ,_+_= 5*= 51:8`= :3= (&8= )*= 343885S(.+.(.9S= &= S9S= 7S;S1S= (-*?= :3*= &:97*= 47(-.)S*= 2>(4-S9S749745-*`= 1:118(-&*,*1.&= &5->11& = 6:.= *89= &884(.S*= A= ).++S7*398= 8&5745->9*8= A= 1&= :&)*14:5*= B_= #&7948= *9= #_8_= *1488*`= (422:3.(&9.43= 5*78433*11*_= &,.98.1= ):3= (&8= 5&79.(:1.*7= 57457*= &:= 2.1.*:= 9745.(&1== '&= 85S(.+.(.9S= *=.89*= *3= 2.1.*:= 9745.(&1`= 9&39= ;.88A8;.8= )*= (&-25.,3438= 8&5745->9*8`= 5&7&8.9*8= 6:*(942>(47-.?.*38= .&'1*&:= -_= 03*= &:97*= *85T(*= )*= 3*499.S*8= 57S8*39*= *3= .-&1&3)*`= *5-&1&39-*7&= *=.,:& `= 8*89=7S;S1S*=&884(.S*=*11*=&:88.=A=)*8=(-&25.,3438=*(942>(47-.?.*38`=)*8=9-S1S5-47&(S*8`= 2&.8=)*=2&3.T7*=51:8=85S(.+.6:*=B.,_=++`= (+ =&79.(1*= _== 1*5*3)&39`= (*99*= *85T(*= 54:88*= 8:7= 1*8= -&:98= 8422*98= )*= .-&1&3)*`= )&38= )*8= +47U98= A= B&,&(S*8= A= .3+1:*3(*= .2&1&>*33*`= *9= 84:;*39= 51:8= -&:9= *3= &19.9:)*= 6:*= 1*8= 5->1147(-.8 _=1*8=+47U98=-:2.)*8=8439=(438.)S7S*8=51:8=574(-*8=)*8=+47U98=9*25S7S*8=6:*= )*8=+47U98=A=).59*74(&75&(S*8=6:.=1*8=*394:7*39=1*=&3)*7=*9='**`=,**/_='*85T(*= _=*=.,:& = 3*=(4389.9:*=)43(=5&8=:3=*=*251*=)*=85S(.+.(.9S=9745.(&1*`=(4259*=9*3:=)*=1*3;.7433*2*39= )&38=1*6:*1=*11*=54:88*_=4:7=(4257*3)7*=1*8=7&.8438=)*=(*99*=343885S(.+.(.9S`=.1=+&:9=)43(= 8.39S7*88*7=A=1&=+4.8=&:==5&79.(:1&7.9S8=):=2.1.*:=9745.(&1`=*9=A=(*11*8=):=,*37*= 5->1147(-.8 _= = -_- 03*=5&79.(:1&7.9S=)*8=+47U98=)8.*=):=:)8*89== = '*8= +47U98= A= ).59*74(&75&(S*8= *9= A= +&,&(S*8= 8439= 9>5.6:*8= )8.*= ):= :)8*89`= *9= ).++S7*39*8= )*8= &:97*8= +47U98= 9745.(&1*8= 343= 8*:1*2*39= 5&7= 1&= )42.3&3(*= )*= (*8= )*:== +&2.11*8=)&7'7*8`=2&.8=&:88.=5&7=1&=).;*78.9S=)*=(-&25.,3438=*(942>(47-.?.*38=57S8*398=a= *3= *++*9`= (*8= )*:== +&2.11*8= )&3,.485*72*8= 8439= &884(.S*8= A= )*8= *85T(*8= *(942>(47-.?.*33*8= 1*=&3)*7`= +320=b= .*)*7844= *9= &1_` = ,**1'=b= ,**1(`= 349&22*39= )*= 7:88:1&(S*8`=)*=9-S1S5-47&(S*8`=)*=8*'&(.3&1*8=*9=)*=(1&;:1.3&(S*8=$.;.T7*= *9=&1_` =,**1_='*8= &:97*8=+47U98=9745.(&1*8=8439=(433:*8=54:7=U97*=51:99=)42.3S*8=5&7=)*8=*85T(*8=)&7'7*8= *3)42>(47-.?S8_=5&38=(*8=+47U98=)8.*`=)*=342'7*:8*8=*85T(*8=)&7'7*8=&884(.S*8=A=)*8= (-&25.,3438= *3)42>(47-.?.*38= 8439= &:88.= 57S8*39*8= #4>*784*3= *9= &1_` = ,**+=b= 2.9-= ]= $*&)`= ,**2_= !3= (438S6:*3(*`= 343= 8*:1*2*39= (*8= +47U98= 8439= -S9S74,T3*8= &:= 3.;*&:= ):= (4:;*79=;S,S9&1`=2&.8=&:88.=&:=3.;*&:=)*=1&=7S5&79.9.43=)*8=2>(47-.?*8=1*=&3)*7=]='**`= ,**/_= I:&39= &:= (479T,*= )*= (-&25.,3438= *(942>(47-.?.*38`= .1= *89= 97T8= 8.2.1&.7*= &:== (479T,*8= 9*25S7S8= *3= 9*72*8= )*= +&2.11*8= )*= (-&25.,3438`= &;*(= :3*= )42.3&3(*= )*8= 7:88:1&(S*8=*9=)*8=9-S1S5-47&(S*8=B.,_=+/_=1*79&.38=&:9*:78=&;&3(*39=6:*=1&=).;*78.9S=)*= (-&25.,3438=*(942>(47-.?.*38=>=*89=51:8=,7&3)*=6:*3=2.1.*:=9*25S7S=1*=&3)*7=]='**`= ,**/= 9&3).8= 6:*= )&:97*8= 84:9.*33*39= 1*= (4397&.7*= 8&7&`= (422:3.(&9.43= 5*78433*11*_= 1*5*3)&39`= 1*8= S9:)*8= )*8(7.59.;*8= 8:7= 1*8= (479T,*8= )*= 2>(47-.?*8= 8439= *3(47*= 7&7*8= *3= (425&7&.843=)*=(*11*8=7S&1.8S*8=*3=2.1.*:=9*25S7S_=5*=51:8=1*8=S9:)*8=8439=54:7=1&=51:5&79= +43)S*8= 8:7= 1*8= (&7545-47*8`= 6:.= 3*= 7*+1T9*39= 5&8= 1&= ).;*78.9S= 7S*11*= )*8= (-&25.,3438= *(942>(47-.?.*38=B.,_=+/`=$.(-&7)= *9=&1_` =,**.= '-S9S74,S3S.9S= 85&9.&1*= )*= 1&= ).897.':9.43= )*8= *85T(*8= 2>(47-.?.*33*8= 54:77&.9= +&;47.8*7= 1&= 343885S(.+.(.9S= *3= 5*72*99&39= )*8= .39*7&(9.438= 2:19.51*8= 7:38`= +33/= *9= *3= 8S1*(9.433&39= )*8= *85T(*8= ,S3S7&1.89*8= (&5&'1*8= )*= 8&884(.*7= A= 51:8.*:78= 5&79*3&.7*8= ).++S7*398= 5&7= 7&55479= A= )*8= *85T(*8= 85S(.&1.89*8= )439= 1*8= 5&79*3&.7*8= )*;.*33*39= 24.38= +7S6:*398_==

50 = = = =

/=(2 = +=(2 =

5->1147(-.8=2439&3& = *5-&1&39-*7&=*=.,:& = = = = B.,:7*=+0_= -494,7&5-.*8= .3=8.9: =)*8=&55&7*.18=;S,S9&9.+8=*9=7*574):(9*:78=&S7.*38=)*= _=2439&3& =*9=)*= _= *=.,:& =.3=8.9:_=

51 5*8=24)T1*8=9-S47.6:*8=(43+479*39=)&.11*:78=(*99*=->549-T8*`=*9=24397*39=6:*=)&38= :3= *3;.7433*2*39= -S9S74,T3*`= 1*8= 897&9S,.*8= ,S3S7&1.89*8= 8439= 51:8= &;&39&,*:8*8= C&3= ..*3)*7*3`=+331=b=$43(*=]=A.705&97.(0`=,**+=b='&3(-.*7=]=8*:-*:8*7`=,**0_=1*8=897&9S,.*8= ,S3S7&1.89*8= 8439= ) &.11*:78= +7S6:*39*8`= 349&22*39= (-*?= 1*8= (-&25.,3438= *(942>(47-.?.*38= (+_ &79.(1*= `= 47943=]=7:38`=+332=b=,**+=b=#41.3&= *9=&1_` =+33,=b=2.9-=]= $*&)`= ,**2_= 5*= 51:8`= 1*8= (-&25.,3438= *(942>(47-.?.*38= 574):.8*39= )*8= 8547*8= 6:.= 8*= ).85*78*39=8:7=)*=143,:*8=).89&3(*8`=*9=U97*=,S3S7&1.89*=5*:9=U97*=:3=&;&39&,*=54:7=,*72*7= )&38= )*8= *3;.7433*2*398= 97T8= ).;*78_= = 1.389&7= )*8= (-&25.,3438`= 1*8= 47(-.)S*8= 574):.8*39= )*= 5*9.9*8= ,7&.3*8`= (&5&'1*8= )*= ).85*78*7= 8:7= )*= 143,:*8= ).89&3(*8`= (*= 6:.= 54:77&.9= +&;47.8*7= :3*= 897&9S,.*= ,S3S7&1.89*_= !++*(9.;*2*39`= )*= 342'7*:8*8= 47(-.)S*8= (-14745->11.*33*8= &55&7&.88*39= 5*:= 85S(.+.6:*8= (+ = 5&7&,7&5-*= +_/= 2&.8= 1&= 51:5&79= )*8= 2>(4-S9S749745-*8=8439=85S(.&1.89*8_=5&38=1&=2*8:7*=41=(*8=)*73.T7*8=8*=(425479*39=51:8= *3= 5&7&8.9*8= *3(&)7S= /= )*8= (-&25.,3438`= (*99*= &884(.&9.43= 54:77&.9= 9*3)7*= A= 51:8= )*= 85S(.+.(.9S= )&57T8= 1->549-T8*= )*= (4S;41:9.43= 5&7&8.9&.7*_= !3= 6:4.= 1*8= 5->1147(-.8 = S(-&55*7&.*398.18= A= 1&= 7T,1*== '*:7= 89&9:9= 8*7&.98.1= =24.38= 5&7&8.9&.7*== 6:*= 1*8= &:97*8= 2>(4-S9S749745-*8== = '&=349.43=)*=(429=)*=1&884(.&9.43=&:=3.;*&:=.3).;.):*1=*89=7*1&9.;*2*39=).++.(.1*=A= S9&'1.7=54:7=(*8=(-&25.,3438`=343=8*:1*2*39=5&7(*=6:.18=8439=).++:8=)&38=1*=841=*9=5&7(*= 6:*=1*8=(4298=5488.'1*=)4.;*39=U97*=(&1(:1S8=A=5&79.7=)*8=+1:==):=(-&25.,343=;*78=1&=51&39*= 2>(4-S9S749745-*= *9= 343= )*= 1&= '.42&88*= +.3&1*= 6:.= 7*57S8*39*= :3*= 6:&39.9S= 51:8= 5*9.9*_='*= )*,7S= )-S9S749745-.*= )*8= 51&39*8= 2.=49745-*8$= 57S8*39*= :3= ,7&).*39`= 41= (*79&.3*8=*85T(*8=7*O4.;*39=,*=A=-*=)*=1*:7=(&7'43*=)*8=(-&25.,3438=.)&7943)4= *9=&1 _`= ,**.=9&3).8=6:*=)&:97*8=*3=7*O4.;*39=1&=6:&8.8949&1.9S= *'&:*7=]=#*>*7`=,**-=b= .71&3)&= *9= &1_ `= ,**0= *9= 8*= 7&5574(-*39= )*8= 51&39*8= 2>(4-S9S749745-*8_= '&= 51:5&79= )*8= 51&39*8= 2.=49745-*8$= 3*= 8439= 5&8= 85S(.+.6:*8`= 8&:+= .24)47:2= &'479.;:2 = 6:.= 7*O4.9= 94:9= 843= (&7'43*= )*8= (-&25.,3438= 94:9= *3=S9&39= 5-4948>39-S9.6:*`= .71&3)&= *9=&1 _`= ,**0_= %7= (*8= 51&39*8= 2.=49745-*8$= 8439= &= 57.47. = 24.38= (429*:8*8= 6:*= 1*8= 51&39*8= 2>(4-S9S749745-*8= 54:7= 1*8= (-&25.,3438_= .= 1*8= 5->1147(-.8 = 8439= :3= 5*:= 24.38= (429*:8*8= 6:*= 1*8= &:97*8= 2>(4-S9S749745-*8`=&1478=(*1&=54:77&.9=2&.39*3.7=1*:7=343885S(.+.(.9S_== = '&884(.&9.43= &:== 5->1147(-.8 = 54:77&.9= U97*= 7*1&9.;*2*39= 24.38= (429*:8*= 343= 5&8= 5&7(*= 6:*= (*8= 2>(4-S9S749745-*8= 7*O4.;*39= 24.38= )*= (&7'43*= )*8= (-&25.,3438`= 2&.8= 5&7(*= 6:*= 1*8= (-&25.,3438= *3= 7*O4.;*39= 51:8= )*8= &7'7*8_= !3= *++*9`= 1&(9.;.9S= 5-4948>39-T9.6:*= *89= 51:8= .25479&39*= *3= 2.1.*:= 9745.(&1= 84:8= 1*++*9= )*8= 51:8= +479*8= 9*25S7&9:7*8= *9= ):3*= 51:8= ,7&3)*= 5S7.4)*= )*= 7S&1.8&9.43= )&38= 1&33S*= 5&8= )-.;*7= 4:= 8&.843= 8T(-*= 7S):.9*`= (*= 6:.= 2T3*= A= :3*= 51:8= ,7&3)*= 6:&39.9S= )*= (&7'43*= &88.2.1S_= %7`= 14786:*= 1&88.2.1&9.43= &:,2*39*`= 1*8= (-&25.,3438= *(942>(47-.?.*38= 7*O4.;*39= 51:8= )*= (&7'43*=1'*7943= *9=&1_` =,**1_=5*=51:8=*3=S9&39=&884(.S*=A=51:8.*:78=(-&25.,3438=A=1&=+4.8`= (-&6:*=(-&25.,343=+4:73.9=.3).;.):*11*2*39=24.38=)*=(&7'43*`=(*=6:.=5*:9=8S1*(9.433*7`= &:=3.;*&:=)*=1*85T(*`=:3*=897&9S,.*=,S3S7&1.89*_= =

52 = = = = &= (=

7= 9=

7- =

/=(2 = +=(2 =

'= )=

7- = 7= 5=

7- =

+=(2 = +=(2 =

= = = B.,:7*=+1 _=>89T2*8=7&(.3&.7*8=) _=2439&3& =&`='=*9=)*= _=*=.,:& =(`=)=a=;:*=)*38*2'1*=*9=)*=)S9&.1_= 7-=a=7-.?42*`=7=a=7&(.3*`=5=a=&3(.*33*=7&(.3*`=&99&6:S*=5&7=)*8=(-&25.,3438=5&9-4,T3*8`=9=a=9.,*= +*:.11S*=5479&39=)*8=+1*:78_=

53 -_. 03*=5&79.(:1&7.9S=)*8= 5->1147(-.8 == = 5&38= 1&= 2*8:7*= 41= )&:97*8= 47(-.)S*8= 2>(4-S9S749745-*8= 8439= &884(.S*8= 85S(.+.6:*2*39=A=)*8=(-&25.,3438=*(942>(47-.?.*38=)&38=(*8=2U2*8=+47U98`=.1=*89=5488.'1*= 6:*= (*99*= 343885S(.+.(.9S= 84.9= :3= (&7&(9T7*= ):= ,*37*= 5->1147(-.8 _= !9433&22*39`= 5->1147(-.8 =&=54:7=,74:5*=+7T7*=1*=,*37*= .24)47:2 `=)*8=2.=49745-*8$=97T8=85S(.+.6:*8=)*8= 7:88:1&(S*8= B.,_= 2_= '*8= 5->1147(-.8 = ).++T7*39= 2475-414,.6:*2*39= )*8= &:97*8= 2>(4-S9S749745-*8=5&7=)*:==&85*(98=a=.=1&=9&.11*=.25479&39*=)*=1*:7=8>89T2*=7&(.3&.7*`=6:.= 54:77&.9= 5*72*997*= )*8= .39*7&(9.438= 2:19.51*8= *9= ..= 1&= -&:9*:7= .25479&39*= )*= 1*:7= 9.,*= +1*:7.*= 6:.= 5479*= *3= ,S3S7&1= &:= 24.38= :3*= ).?&.3*= )*= +1*:78= *9= 5*:9= )43(= 5*72*997*= 1&= ).85*78.43=)*8=,7&.3*8=*3=51:8=,7&3)*=6:&39.9S=*9=A=51:8=,7&3)*=).89&3(*=B.,_=+0=*9=+1`=47= 54:7=)*8=,7&.3*8=6:.=).85*78*39=14.3`=.1=*89=51:8=&;&39&,*:==)U97*=343885S(.+.6:*_= .4:9= )&'47)`= 1&= 51:5&79= )*8= 51&39*8= 2>(4-S9S749745-*8= 57S8*39*39= :3= 8>89T2*= 7&(.3&.7*=(425&(9`=2&.8=*=97U2*2*39=7&2.+.S`=(*=6:.=&:,2*39*=1&=8:7+&(*=)S(-&3,*=&;*(= 1*8= (-&25.,3438_= 1*8= 8>89T2*8= 7&(.3&.7*8= 8439= )&.11*:78= *=97U2*2*39= 24).+.S8= 5&7= 7&55479=A=(*:==)*8=*85T(*8=&:949745-*8='*&0*`=+33._=&7=*=*251* =_=*=.,:& =57S8*39*=:3= 7-.?42*=(4:79=+8,=(2=5479&39=*3=,S3S7&1=)*8=7&(.3*8= &3(.*33*8= 97T8= 2>(47-.?S*8= 2&.8= &:==5*149438=)S,S3S7*8(*398`=143,:*8=&:=2&=.2:2=)*=+*=(2`=*9=)*8=7&(.3*8=51:8=7S(*39*8`= 51:8= (4:79*8= +8,= (2`= 5&7+4.8= 2>(47-.?S*8= *9= 8:794:9= 7.(-*8= *3= &2.)43= B.,_= +0_= '*= ).&2T97*=)*=(*8=7&(.3*8=3*=)S5&88*=/&2&.8=,=22_==145548S`=1*8=7&(.3*8=) 5->1147(-.8 =3*= 8439=5&8=7&2.+.S*8`=8439=51:8=1&7,*8=.=22=*3=24>*33*`=5&79*39=):=7-.?42*=*9=8*3+43(*39= )&38=1*=841=/:86:A=/*=(2=):=7-.?42*=*9=)&38=51:8.*:78=).7*(9.438_='*=;41:2*=*=5147S=5&7= 1*8=7&(.3*8=):3=.3).;.): =&;4.8.3*=84:;*39=1*=2T97*=(:'*=4'8*7;&9.43=5*78433*11*`=B.,_=+1_= !3= (438S6:*3(*`= (*8= 7&(.3*8= 5*:;*39= *397*7= *3= (439&(9= &;*(= 51:8.*:78= *85T(*8= )*= (-&25.,3438`=(422*=1*:7=).897.':9.43=)&38=1*=841=*89=84:;*39=-S9S74,T3*=$.(-&7)= *9=&1_` = ,**/_=5&.11*:78`=14'8*7;&9.43=)*8=7&(.3*8=7S;T1*=)*=,7&3)*8=;&7.&9.438=)*=(4:1*:78=1*=143,= )*=1&=7&(.3*`=8:,,S7&39=1&=57S8*3(*=)*=2>(47-.?*8=&;*(=)*8=54.398=).3+*(9.43=).++S7*398=b= (*8= 2>(47-.?*8= (477*8543)*39= )&57T8= 1.)*39.+.(&9.43= 241S(:1&.7*= A= )*8= (-&25.,3438= ).++S7*398= B.,_= +1_= 1*= 8>89T2*= 7&(.3&.7*= 7*88*2'1*= *3= 94:9= 54.39= A= (*1:.= )*8= *85T(*8= (-14745->11.*33*8= *5-&1&39-*7&=)&2&843.:2 =*9= _=143,.+41.& =6:.=5488T)*39=&:88.=:3=(479T,*= ).;*78.+.S= )*= (-&25.,3438_= .4:8= (*8= S1S2*398= 24397*39= 6:*= 1&= 343885S(.+.(.9S= 5*:9= U97*= (477S1S*=A=:3=8>89T2*=7&(.3&.7*=S9*3):_=5&.11*:78`=54:7=:3*=47(-.)S*=85S(.+.6:*`=.1=3*89= 5&8= 3S(S88&.7*= )S9*3)7*= 843= 8>89T2*= 7&(.3&.7*= :3*= +4.8= 1*= 5&79*3&.7*= 2>(47-.?.*3= 7*3(4397S`=(*=6:.=7*3+47(*=(*99*=(477S1&9.43_== 5&:97*= 5&79`= 1*8= *85T(*8= 2>(4-S9S749745-*8= 439= 84:;*39= )*8= 9.,*8= 7S):.9*8= 6:.= &55&7&.88*39=+:,&(*2*39=1478=)*=1&=+147&.843='*&0*`=+33.`=(422*= _=*=.,:& =)439=1&=9&.11*= 3*=)S5&88*=5&8=+*=(2`=*9=3*=5479*=*3=,S3S7&1=6:*=)*:==+1*:78=B.,_=+0_= _=2439&3& =57S8*39*= )*8=9.,*8=)*=.-=A=+**=(2=)*=-&:9*:7`=6:.=5479*39=*397*=+*=*9=,*=+1*:78=*9= _=(&:)&9& =574):.9= )*8= 9.,*8= 6:.= 5*:;*39= &;4.8.3*7= +`,*= 2= *.)*3+&)*3`= +312_= 1*= 6:.= ).++S7*3(.*= 1*8= 5->1147(-.8 = )*8= &:97*8= (*89= 8:794:9= 1&= -&:9*:7= A= 1&6:*11*= 8439= 574):.9*8= 1*8= +1*:78=a= 1&= -&:9*:7= 7*574):(97.(*_= %7= (*99*= 2*8:7*8= *89= (438.)S7S*= (422*= :3= 97&.9= 6:.= 7*+1T9*= 1*= 54:;4.7= )*= ).85*78.43= ):3*= *85T(*= *3= S(414,.*= +43(9.433*11*= C.411*= *9= &1_` = ,**1_= '*8= 47(-.)S*8=8439=(433:*8=54:7=574):.7*=)*8=,7&.3*8=2.3:8(:1*8=/*8+**= 2=)*=143,=*3=97T8= ,7&3)=342'7*=0,**=5&7=+7:.9`=(-*?= 77(-.8=247.=b =5&7<.3`=+20,`=(*=6:.=1&.88*=8:5548*7=:3*= ).85*78.43= 143,:*= ).89&3(*= )*= (*11*88(._= .= 1*8= 5->1147(-.8 = ).85*78*39= 51:8= 14.3= 6:*= 1*8= &:97*8`=&1478=.1=5*:9=U97*=)&:9&39=51:8=&;&39&,*:==54:7=8*8=,7&.3*8=)U97*=,S3S7&1.89*8_=

54 = = &= 7-.?42* =

7&(.3*8 =

= $:88:1&(S* =K-0 = = = .-S1S5-47&(S *=K+,`=K+- = $:88:1&(S* =K-/ = .-S1S5-47&(S*= =K-, =

= .-S1S5-47&(S* =K+- = $:88:1&(S* =K+- = = $:88:1&(S*=K2`= .-S1S5-47&(S*=K--= +=(2 =

'= 7-.?42* =

7&(.3*8 = .>2342>(*8 = = 2&3.9&(S* =K, = = .-S1S5-47&(S*== = K+*`=K-+ = = $:88:1&(S* =K+, = = 11&;:1.3&(S* =K=0 =

.-S1S5-47&(S*== = = 11&;:1.3&(S* =K+* = K3 = = $:88:1&(S*=K.1= .-S1S5-47&(S*=K3= 11&;:1.3&(S* =K+* = = +=(2 =

= B.,:7*=+2_= $*57S8*39&9.43=8(-S2&9.6:*=)*=1&=7S5&79.9.43=)&38=1*85&(*=)*8=).++S7*398=(-&25.,3438= .)*39.+.S8=A=5&79.7=)*8=7&(.3*8=)*=)*:==.3).;.):8=I+=&=*9=51_+='=) 5->1147(-.8=2439&3&_= '*8=?43*8= -&(-:7S*8=.3).6:*39=6:*=)*8=(-&25.,3438=).++S7*398=574;.*33*39=)*=1&=2U2*=8*(9.43=7&(.3&.7*=+8,= 22=)S5&.88*:7_= = =

55 -_/ '*8=?43*8=)42'7*8=) 5->1147(-.8= = 1*99*= 343885S(.+.(.9S= )*8= 5->1147(-.8 `= 6:.= 7*88*2'1*= A= (*11*= )*= 1&= 51:5&79= )*8= 3S499.S*8= 2.=49745-*8`= *89= 574'&'1*2*39= :3= 97&.9= &3(*897&1= )*= (*= ,*37*= 6:.= 3&= 5&8= S9S= (4397*88S1*(9.433S=)&38=1*8=+47U98=A=).59*74(&75&(S*8=*9=A=+&,&(S*8`=&1478=6:*=)&:97*8=97&.98= 439=(43;*7,S=;*78=(*:==)&:97*8=2>(4-S9S749745-*8`=(422*=1&=7S):(9.43=)*8=+*:.11*8_== 1*5*3)&39`= .1= 7*89*= )&:97*8= 54.398= A= S(1&.7(.7= 8:7= 1*8= 7&(.3*8= ) 5->1147(-.8 _= 1*99*= 343885S(.+.(.9S=*898*11*=+43(9.433*11*==.=94:8=1*8=&7,:2*398=5479*39=A=(74.7*=6:.1=3*=8&,.9= 5&8=)*=(439&2.3&398`=*9=6:.1=*=.89*='.*3=)*8=5*149438=)*85T(*8=).++S7*39*8`=(*8=5*149438= +43(9.433*398.18=94:8=*3=2U2*=9*258==B43(9.433*398.18=94:8==!3=*++*9=(*99*=).;*78.9S=*9=1&= 57S8*3(*= )*= 342'7*:8*8= *85T(*8= 7&7*8= 5*:;*39= 8:,,S7*7= :3*= &((:2:1&9.43= )*= (-&25.,3438= &:= (4:78= ):= 9*258_= 1= *89= 5488.'1*= 6:*= 1*8= 7&(.3*8= /*:3*8= 5488T)*39= )*8= 5*149438= ):3*= 8*:1*=*85T(*`= 97T8=*++.(&(*`=*9=6:*= 1*8=7&(.3*8= B,S*8= 84.*39= &'&3)433S*8= A= )*8= (-&25.,3438= 51:8= ).;*78= *9= 24.38= *++.(&(*8_= 5*8= (&8= )*= 85S(.&1.8&9.43= +43(9.433*11*= ;.88A8;.8= )*= (-&25.,3438= 1478= )*= 1&= ,*72.3&9.43= 439= )S/A= S9S= 2*39.433S8= (-*?= :3*= 47(-.)S*=9745.(&1*=S5.5->9*`= !43458.8=:97.(:1&7.4.)*8 =%9*74= *9=&1_` =,**/=b=,**1=*9=(*1&=3*=(1:*= 5&8=6:*=(*8=47(-.)S*8=8&884(.*39=&:88.=A=)*8=(-&25.,3438=24.38=*++.(&(*8=*9=+472*7=)*8= 5*149438=343=+43(9.433*18_= '&= (425&7&.843= )*8= (479T,*8= )*= (-&25.,3438= .)*39.+.S8= A= 5&79.7= )*8= 7&(.3*8= )*8= 5->1147(-.8 = *9= A= 5&79.7= )*(942>(47-.?*8= )&38= 1*= 347)= )*= 1&= .-&1&3)*= 24397*= 6:*= 1&= 51:5&79= )*8= +&2.11*8= 8439= &884(.S*8= A= 1&= +4.8= &:== 5->1147(-.8 = *9= &:== &7'7*8`= *9= 6:*= 1*8= +&2.11*8=1*8=51:8=+7S6:*39*8=8439=1*8=2U2*8=7:88:1&(S*8`=9-S1S5-47&(S*8`=B.,_=+/_=.*3=6:*= 1*8=S9:)*8=3*=5479*39=5&8=8:7=1*=82U2*8=8.9*8=*9=3*=8439=)43(=5&8=(425&7&'1*8`=:3*=)*8= +&2.11*8=*(942>(47-.?.*33*=1*8=51:8=+7S6:*39*8=(-*?=1*8=&7'7*8`=1*8=8(1*74)*72&9&(S*8`=3&= 54:79&39=5&8=S9S=4'8*7;S*=(-*?=1*8= 5->1147(-.8 =B.,_=+/_=1*99*=&'8*3(*=5*:9=U97*=):*=A=:3= '.&.8= )S(-&39.11433&,*`= 5:.86:*= 1*8= 8.9*8= 8439= ).++S7*398_= 1*5*3)&39`= 1*8= (4:7'*8= )*= 7&7S+&(9.43=8.2:1S*8=54:7=1*8=5->1147(-.8=24397*39=6:*=1*=(479T,*=*89=&88*?='.*3=)S(7.9= (+ _= &79.(1*= =*9=5*:=)*=34:;*11*8=*85T(*8=8439=&99*3):*8=*3=S(-&39.11433&39=51:8=).3).;.):8_= 1*99*=&'8*3(*=)*8=8(1*74)*72&9&(S*8=5*:9=&:88.=7S;S1*7=:3*=8S1*(9.;.9S=)*8= 5->1147(-.8 =*9= )&38=(*=(&8=7S;S1*7=:3*=(*79&.3*=85S(.&1.8&9.43=+43(9.433*11*_== 1=8*7&.9=3S(S88&.7*=)S9:).*7=1&=(422:3&:9S=)*=(-&25.,3438=*(942>(47-.?.*38=A= 574=.2.9S=)*8=7&(.3*8=) 5->1147(-.8 `=54:7=(43+.72*7=4:=.3+.72*7=(*99*=8S1*(9.;.9S`=54:;4.7= (425&7*7=1&'43)&3(*=)*8=).++S7*398=9&=438`=*9=)S(7.7*=1&=897:(9:7*=)*=(*99*=(422:3&:9S`= )439=1-S9S74,S3S.9S=7*89*=->549-S9.6:*_= ==

56 = = &=

'= +***=02 =

45:1&9.438=)*= _=2439&3& = )&38=)*8=+47U98a== 8*25*7;.7*39*8= 8T(-*8=A== (= ).59*74(&75&(S*8= +/*=02 = )*= .382-U3*8=

3).;.):8=)*= _=2439&3& =)439=1&7'7*=1*=51:8= 574(-*=0+2=*89=a=

B&,&( S*8 = &89&3458.8 =85_= .9-4(&75:8 =85_=

.59*74(&75&( S*8 = ,.59*74(&75&8 =85_= /*=2 = #-47*& =85_= = = B.,:7*=+3_= $S5&79.9.43=):=,*37*= 5->1147(-.8 =*3=8.*=)&57T8= &9*2&3= *9=&1_` =,**/ `=)*= _=2439&3&= *3=.-&1&3)*=)&38=).++S7*398=9>5*8=)*=+47U98='`=)42.3S*8=)&38=(-&6:*=7S,.43=5&7=)*8=*85T(*8=)*= +&,&(S*8= 4:= ).59*74(&75&(S*8= ).++S7*39*8`= *9= A= 54.= :9-*5= )&38= 1*= 847)= )*= 1&= .-&1&3)*= (_= '&= (4:1*:7=)*8=(*7(1*8=.3).6:*=1*=9>5*=)*=+47U9=4:=)&7'7*=2>(47-.?.*3=6:.=*394:7*39=1*8=2*39.438=) _= 2439&3& _==

57 .= 1&= 343885S(.+.(.9S= *89= *++*(9.;*2*39= :3= 97&.9= )*8= 5->1147(-.8 `= 43= 5*:9= 8*= )*2&3)*7`= A= 1.389&7= )*= 1&= 85S(.+.(.9S`= 6:*18= *3= 8439= 1*8= &;&39&,*8= *9= 1*8= .3(43;S3.*398= &884(.S8_=1*99*=343885S(.+.(.9S=5*72*9=574'&'1*2*39=&:== 5->1147(-.8 =)*=,*72*7=A=).89&3(*= )*8=5.*)8=5&7*398=*9=)*=(4143.8*7=51:8=)*=+47U98=B.,_=+3_='*8=-&25*8=+147&1*8=)*= _=2439&3& = 439= S9S= *3(148*8= )&38= )*8= 8&(8= *3= 3>143= &;&39= 14:;*79:7*= )*8= +1*:78`= *9= 1&= 51:5&79= )*8= +1*:78=439=3S&324.38=+472S=)*8=+7:.98=)433S*8=343=5:'1.S*8_=1*99*=*85T(*=8*2'1*=)43(= &:94,&2*= *9= )43(= 5*:98U97*= )&:9&39= 51:8= (4143.8&97.(*= )&57T8= 1&= '4.= )*= &0*7= &0*7`= +3//`= 6:.= 89.5:1*= 6:*= 1*8= *85T(*8= &:94,&2*8= 8439= 24.38= (4397*88S1*(9.433S*8= 5&7= 1*8= +&.'1*8=*++*(9.+8=)*=545:1&9.43`=349&22*39=1478=)*=1&=+43)&9.43=):3*=545:1&9.43_== '*8= 545:1&9.438= ) 5->1147(-.8 = 8439= *++*(9.;*2*39= 97T8= ).85*78S*8= A= ).++S7*39*8= S(-*11*8=B.,_=+3=a=A=54.=:9-*5`= _=2439&3& =*89=57S8*39*=)*=342'7*:==8.9*8=A=5&79.7=)*=3**= 2=)&19.9:)*`=84:8=+472*=)*=5*9.9*8=545:1&9.438=).89&39*8=)*=,*=A=+**=2=B.,_=+3_='&.7*=)*= 7S5&79.9.43=)*=(*8=*85T(*8=*89=84:;*39=97T8=S9*3):*=a=.1=*3=*=.89*=--=*85T(*8=).85*78S*8=)&38= 94:9*=18.*=):=:)8*89`=1%(S&3.*=*9=*3=:897&1.*=B.,_=+3_== = '&=,*72.3&9.43=)*8= 5->1147(-.8 =3&=94:/4:78=5&8=S9S=S9:).S*`=47=1&=85S(.+.(.9S=5*:9= ;&7.*7=*397*=1*8=,7&.3*8=*9=1B,*=&):19*=(*=6:.=3*=34:8=5*72*9=5&8=).3+S7*7=1&=85S(.+.(.9S=4:= 1&=343=85S(.+.(.9S=)*8=,7&.3*8=) 5->1147(-.8 _=.=1*8=,7&.3*8=8439=343=85S(.+.6:*8`=&1478=1*:7= ,*72.3&9.43= 5*:9= U97*= +&(.1.9S*= )&38= ).;*78= *3;.7433*2*398_= '&= 1.2.9&9.43= )*= 1&= ,*72.3&9.43=*89=:3=+&(9*:7=)*=(4397*88S1*(9.43=.25479&39=(-*?=1*8=47(-.)S*8_=1*99*=5*79*=)*= ,7&.3*8=6:.=3*=7*3(4397*39=5&8=1*=(-&25.,343=85S(.+.6:*=*89=(438.)S7S*=(422*=:3=(429= &884(.S=A=1&=85S(.+.(.9S=.)&7943)4=]=$*&)`=,**2`=*9=6:.=&=54:7=(438S6:*3(*=)*=1.2.9*7=1*8= *++*(9.+8=)*8=545:1&9.438=)*8=47(-.)S*8_='*8= 5->1147(-.8 =3*=+472*39=5&8=54:7=&:9&39=)*8= 545:1&9.438= )*38*8= *9= .25479&39*8_= 1= *89= 574'&'1*= 6:*= )&:97*8= +&(9*:78= (4397*8 8S1*(9.433*39=(*8=*85T(*8=*9=*3=1.2.9*=1*8=*++*(9.+8`=*9=(*=(429=*89=51:8=574'&'1*2*39=5&>S=A= 1B,*= &):19*= &:= 3.;*&:= .3).;.):*1_= !3= *++*9`= :3*= 4'8*7;&9.43= +7&55&39*= 1478= ):= 57S1T;*2*39= )*8= 7&(.3*8= ) 5->1147(-.8 = *89= 343= 8*:1*2*39= 1&= 9&.11*= )*= 8*8= 7&(.3*8`= 1*8= ;&7.&9.438= )*= (4:1*:78= *9= 1*8= 342'7*:== 54.398= ).3+*(9.438`= 2&.8= 8:794:9= 1*= 342'7*= )*= 7&(.3*8= )S9&(-S*8= ):= 5.*)= 2T7*= *9= 34.7(.*8= 5&7= 1*8= 8&5745->9*8`= *9= 1*= 342'7*= )*= 54.398= ).3+*(9.438= 5&7= )*8= 5&7&8.9*8= 8:7= 1*8= 7&(.3*8= ):= 5.*)= 2T7*= B.,_= +0_= '*8= 7&(.3*8= ) 5->1147(-.8 = *3= 57S8*39*39= '*&:(4:5= 51:8= 6:*= (*11*8= )*= *5-&1&39-*7&= )&2&843.:2= &:=6:*11*8= *11*8= 7*88*2'1*39= 54:79&39_= 1*99*= 343885S(.+.(.9S= *397&3*= 5*:98U97*= :3= 7*1B(-*2*39=)*8=2S(&3.82*8=)*=7*(433&.88&3(*=6:.=1&.88*=1&=;4.*=1.'7*=343=8*:1*2*39=A=)*= 342'7*:== *(942>(47-.?.*38`= 2&.8= &:88.= A= )*= 342'7*:== 5&7&8.9*8= *9= 8&5745->9*8_= 1*99*= 7*1&9.;*= &'8*3(*= )*= 7*(433&.88&3(*= 6:.= 845548*= A= 1&= 8S1*(9.;.9S= 94:9*+4.8= 54:77&.9= *=51.6:*7=1*=+&.9=6:*=)*=342'7*:==8&5745->9*8=&.*39=S9S=&251.+.S8=A=5&79.7=)*8=7&(.3*8=) _= 2439&3& =(+_ &79.(1*= _= = 3= (43(1:8.43`= (*99*= &884(.&9.43= 343885S(.+.6:*= )*8= 5->1147(-.8 = 2S7.9*= :3*= S9:)*=51:8=&5574+43).*`=349&22*39=&:=3.;*&:=)*=1&=,*72.3&9.43`=*9=)*=1&=)>3&2.6:*= 85&9.&1*= *9= 9*2547*11*= )*= 1&= (422:3&:9S= )*= (-&25.,3438= 2>(47-.?.*38_= *8= 5->1147(-.8 =24397*=6:*=1S(414,.*=)*8=2>(4-S9S749745-*8=*89=51:8=).;*78*8=6:*=(*=6:*= 8:,,T7*= 1*= 2.1.*:= 9*25S7S_= 3= 2.1.*:= 9*25S7S`= 8.= 1*8= 2>(4-S9S749745-*8= 439= S9S= 97T8= S9:).S*8`=1*8=2.=49745-*8$=1*=8439=24.38`=*9=3439=S9S=)S(4:7;*79*8=6:*=7S(*22*39_= *= +43(9.433*2*39= )*= (*99*= 3:97.9.43`= *9= 1*8= +&(9*:78= 6:.= 5*:;*39= 8S1*(9.433*7= 4:= (4397*8 8S1*(9.433*7=(*99*=3:97.9.43=2S7.9*39=)U97*=S9:).S8=54:7=2.*:==(4257*3)7*=1S;41:9.43= )*=1&=2>(4-S9S749745-.*_ =

58 =)-S9S749745-.*=*3=(&7'43*= 5&79=)*=(&7'43*=574;*3&39=)*8= (-&25.,3438= 100 80 60 40 20 0

+32*- = *22*. = ,22*. = +22*/ = *12*0 = ,12*0 = &9*8 = = (74.88&3(* = +1 47&.843 = +7:.9 = = = = B.,:7*=,*_= 4:7(*39&,*8=)-S9S749745-.*=*3=(&7'43*=2*8:7S8=8:7=)*8=+*:.11*8=)*= _=)&2&843.:2 =;*798= )*=2&78=A=/:.3=,**1=(&1(:1S=A=(-&6:*=57S1T;*2*39=5&7=7&55479=A=:3*=7S+S7*3(*=&:949745-*` =1*2&9.8= ;.9&1'&` =*9=:3*=7S+S7*3(*=2>(4-S9S749745-*`=)*8=.3).;.):8=&1'.348_= = 85T(*=*9= >(47-.?*8== = =)-S9S749745-.*= &9*= S+S7*3(*= *5-&1&39-*7&= = 2/= ;7.1`=<:.3 = *'&:*7=]=#*>*7`=,**- = )&2&843.:2= !1# = .3= <:.3= <:14:= *9=&1_` =,**/= !1# = --= = .)&7943)4= *9=&1_` =,**.= = !1# = ,*82*= #&788/:.3= 79.(1*== *5-&1&39-*7&=143,.+41.&= !1# = --= <:.3= '&).*= *9=&1_` =,**0= = -2= <:.3= 79.(1*== *5-&1&39-*7&=7:'7&= !1# = ,0= <:.3= *'&:*7=]=#*>*7`=,**- = !1# = 1= = .)&7943)4= *9=&1_` =,**.= 5.5&(9.8=&9747:'*38= !1# = +/= <:.3= *'&:*7=]=#*>*7`=,**- = 5.5&(9.8=-*11*'47.3*= !1# = +.= <:.3= *'&:*7=]=#*>*7`=,**-= = = .*8./= <:.3= 79.(1*== 5.5&(9.8=).89&38= !1# = -0= = .)&7943)4= *9=&1_` =,**.= .89*7&=4;&9&= = ,1= <:.3= *'&:*7=]=#*>*7`=,**-= = = &'1*&:=1_= 4:7(*39&,*=)-S9S749745-.*=*3=(&7'43*=(-*?=1*8=3*499.S*8=2.=49745-*8_= =

59 C_ 5*=1&=2.=49745-.*=A=1&=2>(4-S9S749745-.*= = ._+ '&=2.=49745-.*=*9=843=+43(9.433*2*39= = '*8=51:8=574(-*8=5&7*398=)*8=3*499.S*8=2>(4-S9S749745-*8=439=)S/A=S9S=S9:).S8_=18= 8439= 54:7= 1&= 51:5&79= &884(.S8= *:== &:88.= A= )*8= (-&25.,3438= *(942>(47-.?.*38`= )*= +&O43= 51:8= 4:= 24.38= 85S(.+.6:*= .)&7943)4= *9= &1_`= ,**.=b= *1488*= *9= &1_` = ,**.'=b= .71&3)&= *9= &1_` = ,**0=b= <:14:= *9= &1_` = ,**/=b= '&).*= *9= &1_` = ,**0_= '*:7= 8.,3&9:7*= .84945.6:*= &55&7&9= (422*= .39*72S).&.7*= *397*= (*11*= )*8= 51&39*8= &:949745-*8= 6:.= 1*8= *394:7*39= *9= (*11*= )*8= *85T(*8= 2>(4-S9S749745-*8= 4:= &:97*8= 7S+S7*3(*8= +43,.6:*8=a= (*8= *85T(*8= 8439= 343= 8*:1*2*39= &884(.S*8=A=)*8=(-&25.,3438=*(942>(47-.?.*38=2&.8=*11*8=*3=7*O4.;*39=):=(&7'43*`=(*=6:.= +&.9= )*11*8= )*8= 2.=49745-*8`= A= )*8= )*,7S8= 54:7(*39&,*= )-S9S749745-.*= 51:8= 4:= 24.38= ;&7.&'1*8=.&'1*&:=/`=*1488*= *9=&1_` =,**0_=1*=24)*=)*=3:97.9.43=&=S9S=)S(4:;*79=7S(*22*39= (-*?= 1*8= 47(-.)S*8`= &.38.= 6:*= (-*?= )*8= *7.(&(S*8= 574(-*= )*8= 24349745*8=6:*= 8439= 1*8= 5>741*8=.*)*7844= *9=&1_` =,**1&=b=Z.22*7= *9=&1_` =,**1_=3.9.&1*2*39=S9:).S=(-*?=1*8=3*499.S*8= *'&:*7= ]= #*>*7`= ,**-=b= .)&7943)4= *9= &1_` = ,**.`= 1&= ;&7.S9S= )*8= 24)T1*8= '.414,.6:*8= :9.1.8S8= 54:7= (*8= 7*(-*7(-*8= &= 24397S= 6:*= )*8= *85T(*8= 343= &++.1.S*8= A= )*8= *85T(*8= 2>(4-S9S749745-*8= 5*:;*39= &:88.= U97*= 2.=49745-*8$= = F-.97.),*= ]= 4:9-<479-`= ,**/=b= Z.22*7= *9=&1_` =,**1=b= (+ =&79.(1*= =*9=6:*=(*8=*85T(*8=5*:;*39=U97*=9745.(&1*8= (+_ &79.(1*= _= 1*=9>5*=)*=3:97.9.43=.384:5O433S=.1=>=&=(.36=&38=*3(47*=84:1T;*=)*=342'7*:8*8=6:*89.438= 6:&39= &:= +43(9.433*2*39= )*8= S(-&3,*8= (-&25.,343847(-.)S*_= 5&38= 1&= 2*8:7*= 41= (*8= 47(-.)S*8= 8439= 5-4948>39-S9.6:*8`= 1&= 6:*89.43= *89= )*= 8&;4.7= 8.= *11*8= 8439= (&5&'1*8= )*= )433*7=:3*=5&79.*=)*=1*:7=(&7'43*=47,&3.6:*=&:==(-&25.,3438`=(422*=1&=51:5&79=1*=+439= 14786:*11*8=8439=&884(.S*8=A=)*8=7-.?4(943.&8$=4:='.*3=(4389.9:*398*11*8=94:/4:78=:3=5:.9=)*= (&7'43*== '&= (425&7&.843= )*8= ;&1*:78= )*= 54:7(*39&,*= )-S9S749745-.*= 54:7= _= )&2&843.:2 `=:3*=)*8=2.=49745-*8$=1*8=51:8=S9:).S*8`=24397*=)*=,7&3)*8=;&7.&9.438=*397*= 8.9*8=*9=*397*=)&9*8=.&'1*&:=1b=84398*11*8=):*8=&:==(43).9.438=14(&1*8=)S(1&.7*2*39`=A=1&= 5-S3414,.*= )*8= 51&39*8`= A= )*8= (-&25.,3438= )*85T(*8= ).++S7*39*8`= A= )*8= ;&7.&9.438= ):= +43(9.433*2*39=)*=1&884(.&9.43=*397*=1*8=47(-.)S*8=2.=49745-*8$=*9=1*:78=(-&25.,3438== = *=54:7(*39&,*=)-S9S749745-.*=;&7.*898.1=)&38=1*=9*258== .4:9= )&'47)`= (*8= ;&7.&9.438= *397*= 8.9*8= 3*= 8*2'1*39= 5&8= 7*1.S*8= A= 1&= 1:2.T7*= *'&:*7`=,**/=b=Z.22*7= *9=&1 _`=,**1_== '*8=;&7.&9.438=9*2547*11*8=):=54:7(*39&,*=)-S9S749745-.*=439=S9S=S9:).S*8=(-*?= _= )&2&843.:2` =A=5&79.7=)*8=+*:.11*8=.88:*8=)*=545:1&9.438=3&9:7*11*8= (+_= &79.(1*= _='*:7= +- 1= ).2.3:*=)*=.`=84.9=:3*=;&7.&9.43=)*:==A=6:&97*=+4.8=51:8=,7&3)*=6:*=(*11*=&997.':S*=A=1&= 5-S3414,.*= )*8= +*:.11*8= (-*?= )&:97*8= *85T(*8= 2*)1*>= *9= &1_` = +33+`= 349&22*39= (-*?= 1*2&9.8=;.9&1'&= 8:7=1*=2U2*=8.9*`=)439=1*= +- 1=).2.3:*=)*=,=&:=(4:78=):=57.39*258= (+ _= &79.(1*= _= 5*8= ;&7.&9.438= ):= +- 1= )*= 147)7*= )*= .= 8439= 2*8:7S*8= (-*?= )*8= &7'7*8= &:= 3.;*&:=)*8=9.,*8=*9=343=)*8=+*:.11*8=5&2*8.3=]='*1&7,*`=,**-=b=#&:34:7>= *9=&1_` =,**1_=%7= 1*8=9.,*8=343=5-4948>39-S9.6:*8=439=:3*=8.,3&9:7*=.39*72S).&.7*=*397*=1*8=+*:.11*8=8:(7*8= .88:8= )*= 1&= 5-4948>39-T8*= *9= 1*8= 7&(.3*8= 8:(7*8= .88:8= )*= 1&2.)43`= 6:.= &= :3*= 8.,3&9:7*= *37.(-.*=*3= +- 1`=7:,341.= *9=&1 _`=+322=b=5&2*8.3=]='*1&7,*`=,**-`=*9=(422*=*11*8=:9.1.8*39= )*:==84:7(*8=)*=8:(7*8`=1*8=;&7.&9.438=)*= +- 1=8439=51:8=,7&3)*8=6:*=(*11*8=)*8=+*:.11*8=6:.= 3439=1&=8.,3&9:7*=6:*=):=(&7'43*=.88:=)*=1&=5-4948>39-T8*_='*8=;&7.&9.438=2*8:7S*8=(-*?= _=)&2&843.:2 =8439=)43(=)&:9&39=51:8=8.,3.+.(&9.;*8=6:*11*8=(43(*73*39=1*8=+*:.11*8_=

60 = = = = =

=)-S9S749745-.*=*3=(&7'43* =

+,*+_,= &= &= +**==)-S9S749745-.* = &= &` +** =+ &= (= 2* *_2= a a &= &= '= a &= (= &= (= '= '= (= 0* *_0 '= = '= '= &= .* *_.= &= ,* *_,= *= * BB7 $$-. BB7 $$-. BB7 $$-. BB7 $$-. BB7 $$-.B1 =_=)_ = _=1_ = _=-_ = _=-_ = _=)_ = _=7_ = #439+*77.*7 = B439&7*9 = 1-&:7.&9 = .9*8`=*85T(*8=*9=47,& 3*8 =

$&(.3* =$ = $-.?42*=$-= ..,*=.= B*:.11*8=B= B7:.98 =B7 =

= = = = B.,:7*= ,+_= C&7.&9.438= .39*7847,&3*8= ):= 54:7(*39&,*= )-S9S749745-.*= *3= (&7'43*= )*= .= *85T(*8= )*= 3*499.S*8= a= _= )&2&843.:2 = _= ) _`= _= 143,.+41.& = _= 1 _`= _= 7:'7& = _= 7 _= *9= _= -*11*'47.3* = _= - _= A= #439+*77.*7`=1-&:7.&9=*9=)&38=1*='4.8=):=B439&7*9_='*8=;&1*:78=):=54:7(*39&,*=)-S9S749745-.*=3*= 8439=5&8=).++S7*39.&'1*8=*397*=47,&3*8=14786:*=1*8=1*997*8=&:=)*88:8=)*8='&77*8=8439=.)*39.6:*8=54:7= :3*=*85T(*=*9=:3=8.9*=)433S=9*89=)*=#&338F-.93*>=)*:==A=)*:=`=0*`*/_=

61 '4786:*=1*8=9.,*8=+*:.11S*8=8479*39=)*=9*77*`=1*:7=54:7(*39&,*=)-S9S749745-.*=*3=(&7'43*= 8S1T;*=A=2*=9&3).8=6:*3=+.3=)*=8&.843`=14786:*=1*8=+7:.98=(422*3(*39=A=8S(-*7`=+.3=/:.3`=(*= 54:7(*39&,*=&;4.8.3*=1*8=,*=B.,_=,+_='&=).2.3:9.43=)*=(*=54:7(*39&,*=*89=(439.3:*=*397*= 2&78=*9=/:.3=*9=24397*=6:*=51&39*=&(6:.*79=&:=(4:78=)*=1&=8&.843=:3*=5745479.43=)*=51:8=*3= 51:8=.25479&39*=)*=843=(&7'43*=A=97&;*78=1&=5-4948>39-T8*_='*8=54:7(*39&,*8=3.3).6:*39= 5&8= )*8= 6:&39.9S8= *9= 1*++.(&(.9S= )*= 1&= 5-4948>39-T8*= S9&39= 8:5S7.*:7*= 54:7= 1*8= 51&39*8= 2&9:7*8`=(*8=;&7.&9.438=54:77&.*39=U97*=&997.':S*8=A=1&=8*:1*=&:,2*39&9.43=)*=1&88.2.1&9.43= 5-4948>39-S9.6:*= *7*.7&`= +33/=b= 5&2*8.3= ]= '*1&7,*`= ,**-=b= #&:34:7>= *9= &1_` = ,**1_= 1422*=1*8=;&1*:78=)*= +- 1=*9=8*8=;&7.&9.438=8439=).++S7*39*8=)*=(*11*8=)*= 1*2&9.8=;.9&1'& =4:= -:'.&= 5*7*,7.3& `= )439= 1*8= ;&7.&9.438= )*= +- 1= 8439= :3.6:*2*39= ):*8= &:== (-&3,*2*398= )*= 5-S3414,.*`= )*= 9*11*8= ;&7.&9.438= )*= +- 1= 5*:;*39= )43(= U97*= &997.':S*8= &:88.= A= :3= (-&3,*2*39=)*=3:97.9.43=(&7'43S*_=5*=51:8`= 1*=7&55479=128=&:,2*39*=A=5&79.7=)*=2&._= 1*99*=&:,2*39&9.43=5*:9=&:88.=7*+1S9*7=:3=(-&3,*2*39=)*=3:97.9.43`=*9=349&22*39= :3*= ).2.3:9.43= 7*1&9.;*= )*8= &554798= )&?49*= (425&7S= &:= (&7'43*_= '&?49*= S9&39= &55479S= 5&7= 1*8= 2>(47-.?*8`= 1*8= &554798= *3= 574;*3&3(*= )*8= (-&25.,3438= *&:`= 8*18= 2.3S7&:== *9= (&7'43*= 47,&3.6:*= 8*7&.*39= .25479&398= &;&39= 1&= 2.82&.= +147&.843= *9= ).2.3:*7&.*39=A=5&79.7=)*=2.82&._= = *=54:7(*39&,*=)-S9S749745-.*=;&7.*898.1=*397*=47,&3*8== 5&38=1&=5-S3414,.*=)*= _=)&2&843.:2 `=2.82&.=(477*8543)=A=1&=5S7.4)*=)*=+147&.843= B.,_=,+`=1*8=+7:.98=8439=7&5.)*2*39=+472S8=*9=).85*78*39=1*:78=,7&.3*8=*3=/:.11*9= (+_ &79.(1*= _=!397*=(*8=)*:==5S7.4)*8`=43=5*:9=8*=)*2&3)*7=1&6:*11*=)*8=)*:==84:7(*8=)*=(&7'43*= *89=:9.1.8S*=54:7=+472*7=)*8=47,&3*8=9*18=6:*=1&=+1*:7=4:=1*=+7:.9_=1*99*=S9:)*=)*8=;&7.&9.438= .39*7847,&3*8=):= +- 1=&=57.3(.5&1*2*39=5479S=8:7=1*=+7:.9`=(-*?=51:8.*:78=*85T(*8=B.,_=,*_= '*8= ;&7.&9.438= )*= +- 1= 8439= )*= 147)7*= )*= .= 84.9= )*= 34:;*&:= )*:== +4.8= 51:8= 6:*= 1*8= ;&7.&9.438=*397*=47,&3*8=2*8:7S*8=(-*?=)&:97*8=51&39*8=&)*(0= *9=&1_` =,**/_=5*8=+1*:78=)*= _=7:'7& =439=S9S=S(-&39.11433S*8=*9=439=:3*=8.,3&9:7*=*3= +- 1=343=).++S7*3(.&'1*=)*8=7&(.3*8`= *9= 7*(*;7&.*39= 3*= )*= 1*:7= (&7'43*= )*8= (-&25.,3438`= (4397*= 2*= )&38= 1*8= 9.,*8_= '&= 8.,3&9:7*=.84945.6:*=)*=1*:78=+*:.11*8=3&=5&8=S9S=2*8:7S*=*9=1.2.9*=1*8=(43(1:8.438`=2&.8=1*= 54:7(*39&,*= )-S9S749745-.*= *89= 3S&324.38= 97T8= S1*;S_= !3= 7*;&3(-*`= 1&= 8.,3&9:7*= .84945.6:*=)*8=+7:.98=)*8=974.8=&:97*8=*85T(*8=*89=94:/4:78=.39*72S).&.7*=*397*=1*8=+*:.11*8=*9= 1*8=7&(.3*8`=84:;*39=343=).++S7*39.&'1*=4:=.3+S7.*:7*=A=(*11*=)*8=9.,*8=8=84.9=*397*=./=*9=/*= B.,_=,+=*9=,-`= (+ =&79.(1*= _='*8=+7:.98=&55&7&.88*39=)43(=51:8=&:949745-*8=6:*=1*8=&:97*8= 47,&3*8=&S7.*38=-S9S749745-*8=9.,*`=*9=1*=(&7'43*=.88:=)*=1&=5-4948>39-T8*=5&79.(.5*=)43(= 57S+S7*39.*11*2*39=A=8&=+472&9.43_=1*5*3)&39=8&=8.,3&9:7*=3*89=5&8=.)*39.6:*=&:==+*:.11*8= *9=.1=*89=)43(=(42548S=&:88.=)*=(&7'43*=+43,.6:*_= !3=7&88*2'1&39=1*8=)433S*8=)*8=)*:==2*8:7*8`=8:7=1*8=+*:.11*8=&:=(4:78=)*=1&=8&.843`= *9=8:7=1*8=+7:.98=*3=+.3=)*=8&.843`=43=5*:9=*3=)S):.7*=6:*=1&=2.=49745-.*=;&7.*=&:=(4:78=):= 9*258=*9=&:=8*.3=)*=147,&3.82*_=!11*=*89=2&=.2&1*=*3=)S':9=)*=(74.88&3(*`=14786:*=1&=9.,*= 8479=)*=9*77*=*9=574):.9=1*8=+1*:78`=*9=1&=5-4948>39-T8*=8>=&/4:9*=*38:.9*_=1*99*=:9.1.8&9.43= )*8= 7*884:7(*8= 2.(74'.*33*8= 2&=.2&1*= *3= 5S7.4)*= )*= (74.88&3(*= *89= :3= 5-S342T3*= (4:7&39`=(-*?=1*8=51&39:1*8=)&7'7*8=&884(.S8=A=)*8=(-&25.,3438=*(942>(47-.?.*38=.2&7)= *9=&1_`= +331`=4:=1*8=+&'&(S*8=&884(.S*8=A=)*8='&(9S7.*8=+.=&97.(*8=)&?49*=94:,&&7)`=,***_= 1:8= ,S3S7&1*2*39`= 1:9.1.8&9.43= )*= 7S8*7;*8= 84:9*77&.3*= *3= )S':9= )*= (74.88&3(*= *9= 1.25479&3(*=51:8=9&7).;*=)*=1&=5-4948>39-T8*=*89=:3=5-S342T3*=(4:7&39=5&2*8.3= *9=&1 _`= +332=b=5&2*8.3=]='*1&7,*`=,**-=b=#&34:7>= *9=&1_` =,**1_=

62 = = &=

' 4>*33*=54:7=1*85T(*

1-14745->11 *8=

39-4(>&3*8=

_=&9747:'*38 &1'.+147& 5&11*38 5&11*38 ;.).7.+147& 1:9*8(*38 &1'.+147&

_=5:75:7&9& 748*&= *7)3*7.

_=-*11*'47.3* x x +1&;*8(*38 +1&;*8(*38 (-147&39-&= &1'.+147&= ;.7.)&38 &1'&

= B.,:7*= ,,_= &= 5*= ,&:(-*= A= )74.9*=a= 9:*7(:8= 7:'*7 = /*:3*= &1'.348= 5-494= _= *4++74>`= 5.5&(9.8= -*11*'47.3* =&1'.348=8&38=(-14745>11*=&;*(=&39-4(>&3*8='`=5-494=_=4:1.S=4:=8&38=&39-4(>&3*=(`= 5-494= <_81_= 11&*88*38= *9= '= 5.;*78.9S= )*8= (&8= )*= )S5.,2*39&9.43= (-*?= 5.5&(9.8= &9747:'*38 `= _5:75:7&9& `= _=-*11*'47.3*= )&57T8=A1*.3`=+313=*9=5*1+47,*`=+332_ =

63 ._, '&55&7.9.43=*9=1*=2&.39.*3=)*8=&1'.348=)*8=3*499.S*8== = '&=2.=49745-.*=)*8=47(-.)S*8=&=S9S=)S24397S*=7S(*22*39`=,7B(*=&:==2*8:7*8=)*= +- 1= *9= +/ 8`= 2&.8= (*79&.38= .3).(*8= &;&.*39= 2*3S= A= +&.7*= (*8= 2*8:7*8_= '&= 51:5&79= )*8= 545:1&9.438= )*= *5-&1&39-*7&= *9= ) 5.5&(9.8 = 57S8*39*39= 5&7+4.8= )*8= .3).;.):8= 343= (-14745->11.*38`= ).98= (-14749.6:*8`= ->54(-742*8= *9= 6:*= 34:8= 6:&1.+.*7438= )&1'.348= B.,_= ,,_=18=(477*8543)*39=A=)*8=.3).;.):8=)439=1*38*2'1*=)*8=5&79.*8=&S7.*33*8=+1*:78`=9.,*8= *9=+*:.11*8=8439='1&3(-*8=4:=/&:3*=5B1*=(-*?=1*8= *5-&1&39-*7& =4:=748S8=(-*?=1*8= 5.5&(9.8= 6:.= (439.*33*39= '*&:(4:5= )&39-4(>&3*8= B.,_= ,,_= 1= 3*= 8&,.9= 5&8= ).3).;.):8= )439= 8*:1*= 1&= +1*:7=*89=)S5.,2*39S*_=18=8439=94:/4:78=24.38=+7S6:*398=6:*=1*8=.3).;.):8=;*798=&:=8*.3=)*8= 545:1&9.438`= *9= 8439= 57S8*398= )&38= 5*:= )*= 545:1&9.438_= '*8= )433S*8= 8439= +7&,2*39&.7*8= 2&.8=43=2*39.433*=1&=57S8*3(*=)&1'.348=)&38=0=*85T(*8=)*= *5-&1&39-*7& =*9=+*=) 5.5&(9.8 `= *3= 2S7.6:*= ):= 847)`= )&38= 94:9*= 1!:745*= *9= &:= <&543= .&'1*&:= 0_= := (4397&.7*= )*8= .3).;.):8= &1'.348= 7*3(4397S8= (-*?= )&:97*8= *85T(*8`= (422*= (-*?= 9:*7(:8= 7:'*7 = B.,_= ,,`= 6:.= 3*= 8:7;.;*39= 5&8`= (*8= &1'.348= 8439= )S5.,2*39S8= 94:9*= 1*:7= ;.*`= 8:7;.;*39= *9= 5*:;*39= 574):.7*=)*8=+7:.98= = 85T(*= &>8=*9=342'7*8=)*=2*39.438= 3= *5-&1&39-*7&=)&2&843.:2= B7&3(*`=9&1.*`=11*2&,3*`= `=14;&6:.*= +1+= _=143,.+41.&= B7&3(*`=!85&,3*`=!8943.*= +0= _=7:'7&= B7&3(*= 0= _=0:7).(&= .:76:.*= ,= _=+&1(&9&= <&543= += _=143,.'7&(9*&9&= <&543= += 5.5&(9.8=-*11*'47.3*= B7&3(*`=*1,.6:*`=B.31&3)*`=0= +-*= _=5:75:7&9&= B7&3(*`=11*2&,3*`= = -1= _=2.(745->11&= B7&3(*= 2= _=&1'.*38.8= 414,3*= /= _=97*2418..= B7&3(*= .= _=+.'7..= B7&3(*= -= _=&9747:'*38= B7&3(*`=$S5:'1.6:*=.(-T6:*= -= _=).89&38= B7&3(*= += _=2:*11*7.= 11*2&,3*= += _=3**71&3).(&= &>88&8= += _=1*594(-.1&= 11*2&,3*= += = &'1*&:=2_=42'7*= )*=(&8=)&1'.3.82*=)4(:2*39S8=(-*?=1*8=,*37*8= *5-&1&39-*7& =*9= 5.5&(9.8 =5&7= *85T(*= *9= 5&7= 5&>8= #&.741)= ]= F*'*7`= +3/*=b= &12.&`= +320=b= 5*1+47,*`= +332=b= (&55&9.((.= ]= (&55&9.((.`=+332_=1*8=)433S*8=8439='.&.8S*8=5&7=1*=342'7*=.25479&39=)*=)433S*8=+7&3O&.8*8`=6:.= 8439=.88:*8=)*=2*8=(439&(98=&;*(=1*=7S8*&:=+7&3O&.8=)47(-.)45-.1*8=&2&9*:78_= = 1*=5-S342T3*=&=143,9*258=S9S=&997.':S=A=)*8=).8+43(9.433*2*39=):8=A=1&=57S8*3(*= )*=(-&25.,3438=*3)45->9*8=)*8=+*:.11*8=&12.&`=+320=*9=(*=3*89=6:*=7S(*22*39=6:*=1*:7= 24)*=)*=3:97.9.43=&=S9S=S1:(.)S=a=(*8=.3).;.):8=8439=2>(4-S9S749745-*8=*9=8439=&884(.S8=A= )*8= (-&25.,3438= *(942>(47-.?.*38= 6:.= 1*:78= +4:73.88*39= 343= 8*:1*2*39= )*= 1*&:= *9= )*8= 8*18=2.3S7&:==2&.8=&:88.=):=(&7'43*=*9=)*=1&?49*=47,&3.6:*=*1488*= *9=&1_` =,**.'=b=<:14:= *9= &1_` = ,**/=b= '&).*= *9= &1_` = ,**0_= 1*8= .3).;.):8= 7*88*2'1*39= '*&:(4:5= &:== *85T(*8= (4251T9*2*39=2>(4-S9S749745-*8=)439=.18=8439=574(-*8=B.,_=2_==

64 &

4:7(* 39&,*=) .3).;.):8=*3=+1*:7=4:=*3=+7:.9 0.45 0.4. B147&.843=)*8= 0.35 .3).;.):8=;*798= B7:(9.+.(&9.43=)*8= 0.3- 0.25 .3).;.):8=;*798= B147&.843=)*8= 0.2 , .3).;.):8=&1'.348= 0.15 B7:(9.+.(&9.43=)*8= 0.1+ .3).;.):8=&1'.348 0.05 =0 19/03/2007 08/04/2007 28/04/2007 18/05/2007 07/06/2007 27/06/2007 17/07/2007 &9*8 = '

+=(2

(

/=2

= B.,:7*=,-_= &=-S3414,.*=)*=1&=7*574):(9.43=)*8=.3).;.):8=;*798=*9=&1'.348=)*= _=)&2&843.:2 `='= 574=.2.9S=)*8=)*:==5-S349>5*8=A=#439+*77.*7=*9=(= 7S5&79.9.43= 85&9.&1*= A= #439+*77.*7= *9= 2.8*= *3= S;.)*3(*=)*=1&=7S5&79.9.43=343=&,7S,S*=)*8=5-S349>5*8=)&38=1*85&(*=)_=-494=#_=$4>=

65 97&382.88.43 = 8S1*(9.43 55&7.9.43=*9=2&.39.*3=&:=3.;*&:=)*=1*85T(* = )S7.;*= 4)T1S=+_ = *3;.7433*2*39=8*:1=

2:9&9.43 = 4)T1*=, =a= 2:9&9.43 =8*:1* = 4)T1*=- = 7*574):(9.43 = + 2:9&9.43 = (7>594541>2475-.82* = 2:9&9.43 = 841*2*39=7*574):(9*:7 = 2:9&9.43 = 4)T1*=.=a = 7*574):(9.43 =8*:1* = = B.,:7*= ,._= B&(9*:78= 5*72*99&39= 1&55&7.9.43= *9= 1*= 2&.39.*3= )*8= &1'.348= *9= ).++S7*398= 24)T1*8= 5488.'1*8= 54:7= *=51.6:*7= 1*:7= &55&7.9.43= 7S(:77*39*_= !3= /&:3*`= )*:== ->549-T8*8= (43(*73&39= 1*:7= 89&9:9=54:7=1*85T(*=2.=49745-*_= = 1*8= .3).;.):8= &1'.348= 439= :3*= 8S6:*3(*= .= .)*39.6:*= 4:= 97T8= 5*:= ).++S7*39*= )*= (*:== )*8= .3).;.):8= ;*798= 6:.= 1*8= *394:7*39`= *9= &55&79.*33*39= )43(= A= 1&= 2U2*= *85T(*= )433S*8=343=5:'1.S*8_=+.3=)*=7&.8433*7=8:7=1S;41:9.43=):=(&7&(9T7*=&1'.348=51:8.*:78= 6:*89.438=8439=A=S1:(.)*7=B.,_=,.=a= +_ 6:*1=*89=1*=)S9*72.3.82*=):=(&7&(9T7*=&1'.348==(*=(&7&(9T7*=*898.1=97&382.88.'1*== ,_ (*8= &1'.348= 5*:;*398.18= 8*= 7*574):.7*== 84398.18= 8S1*(9.433S8= 4:= (4397*8 8S1*(9.433S8== -_ (422*39=84398.18=2&.39*3:8==5&7=1&=2:9&9.43`=1*:7=7*574):(9.43=4:=1*8=)*:=== ._ 6:*=7*57S8*39*398.18=54:7=1*85T(*==:3=(7>594541>2475-.82*=&11T1*8=7S(*88.+8= (&(-S8=(-*?=1*8=-S9S74?>,49*8=4:=)*8=545:1&9.438=*3=;4.*=).3).;.):&1.8&9.43== = *=)S9*72.3.82*=)*=1&1'.3.82*=(-*?=1*8=3*499.S*8= 4:7= .)*39.+.*7= 1*= )S9*72.3.82*= )*= 1&1'.3.82*`= .1= +&:)7&.9= .= 97&3851&39*7= )*8= .3).;.):8`= (*= 6:.= *89= .25488.'1*= 54:7= (*8= *85T(*8= &04;80.`= +30/= *9= ..= 54:;4.7= +&.7*= ,*72*7= 1&= )*8(*3)&3(*= .88:*= )*= (74.8*2*398= (43971S8`= (*= 6:.= 5*:9= 3S(S88.9*7= /= &38= $&82:88*3`= +33/_= 8S&324.38`= 1*8= .3).;.):8= &1'.348= *9= ;*798= 8439= 2S1&3,S8= )&38= 1*8= 545:1&9.438=*9=)*8=9.,*8=)*=9>5*8=).++S7*398=54:88*39=5&7+4.8=A=,=(2=)*=).89&3(*=B.,_=,-= 8&38=U97*=(433*(9S*8=&:=2U2*=7-.?42*_=.=1&1'.3.82*=*89=)S9*72.3S=5&7=1*3;.7433*2*39`= (*= 8*7&.9= 84:8= 1.3+1:*3(*= )*= ;&7.&9.438= *=97U2*2*39= 14(&1.8S*8= ):= 841= *9= 343= )*= 1*3;.7433*2*39=1:2.3*:==3.=)*8=*85T(*8=)&7'7*8=4:=)*=(-&25.,3438=&884(.S8_=5*=51:8`= )&38= 1*= (&8= 4:= )*8= .3).;.):8= 8439= &1'.348= .18= 1*= 7*89*39= ):3*= &33S*= 8:7= 1&:97*`= (*(.= .251.6:*7&.9= :3*= 89&'.1.9S= 97T8= +479*= ):= 2.(74*3;.7433*2*39_= *:18= (*79&.38= !5.5&(9.8= -*11*'47.3* =8*2'1*39=&;4.7=:3=(&7&(9T7*=&1'.348=51:8=.389&'1*`=2&.8=3*=7*57S8*39*39=5&8=1&= 2&/47.9S= )*8= (&8_= *>71*= *9= 2.9-= +33-= 439= 24397S= 6:*= )*8= 51&39:1*8= ) 77(-.8= 247.4 = 7*89&.*39= 343= (-14745->11.*38= 8.= .18= S9&.*39= (:19.;S8= 8:7= :3= 2.1.*:= 7.(-*= *3= 8:(7*`= *9= 8:5548&.*39=6:*=1&1'.3.82*=)*8=3*499.S*8=S9&.9=)2=A=:3=&55479=8:551S2*39&.7*=)*=(&7'43*= )*=1&=5&79=)*8=(-&25.,3438_=1*5*3)&39`=1*:7=*=5S7.*3(*=S9&.9=)*=(4:79*=):7S*=*9=.18=3439= 5&8=)S24397S=1&=89&'.1.9S=)*=(*=5-S342T3*=(-*?=(*99*=*85T(*=6:.=3*89=/&2&.8=&1'.348=)&38= 1&= 3&9:7*`= 9&3).8= 6:*= 1*8= &1'.348= )*8= 3*499.S*8= 8439= 89&'1*8= 8:7= 51:8.*:78= &33S*8_= '*8= (43(1:8.438=8:7= 7_=247.4 =3*=34:8=5&7&.88*39=)43(=5&8=97&38548&'1*8=&:==3*499.S*8_=

66 PopC59 PopC43 PopC42 B.,:7*= ,/_= -*34,7&22*= 4'9*3:= A= 5&79.7= ):= PopC31 PopC30 PopC54 PopC58 541>2475-.82*= ,S3S9.6:*= B'= )*= 1&= PopC47 PopC50 PopC46 545:1&9.43= )*= #439+*77.*7= )*= _= )&2&843.:2 = (+ _= PopC53 PopC56 PopC57 &79.(1*= _= '*8= 743)8= '1&3(8= (477*8543)*39= &:== PopC38 PopC36 PopC37 PopC44 .3).;.):8= &1'.348`= 1*8= '7&3(-*8= 8&38= 743)= A= 94:8= PopC52 PopC48 PopC55 1*8=&:97*8=.)3.;.):8=;*798_= PopC35 PopC33 PopC34 PopC28 PopC32 PopC29 = PopC24 PopC25 PopC40 = PopC26 PopC27 PopC41 PopC51 = PopC12 PopC21 PopC23 PopC16 = PopC22 PopC20 PopC14 PopC18 = PopC15 PopC17 PopC7 PopC8 = PopC6 PopC11 PopC3 = PopC13 PopC1 PopC2 PopC4 = PopC5 PopC10 PopC19 PopC39 = PopC45 PopC49 0.1 = _= = = 3(&)7S=0_= *=7*,.2*=7*574):(9*:7=)*= _=)&2&843.:2 = 842'7*:==8439=1*8=&:9*:78=6:.=2*39.433*39=1&:94,&2.*=)*= _=)&2&843.:2 `=/:89.+.S*=5&7=1&'8*3(*= )*=5411.3.8&9*:7=C&3=)*7=1.3,*1`=+33/=b=4:73S7.&8=*9=&1 _`=,**/=*9=1S9:)*=)*=1*:7=897:(9:7*=,S3S9.6:*= (&((-.`= +33*_= 5&57T8= 1 &79.(1*= `= 1&= 897:(9:7*= ,S3S9.6:*= )*= _= )&2&843.:2 = *89= 1&= 2U2*= 6:*= _= 143,.+41.& `=:3*=47(-.)S*=(438.)S7S*=(422*=&114,&2*=(&((-.`=+33*=b=C&3=)*7=1.3,*1`=+33/`=(*=6:.= (4397*).9=1*8=S9:)*8=57S(S)*39*8_=4:7=2.*:==(433&97*=1*=7S,.2*=7*574):(9*:7=)*= _=)&2&843.:2 `= /&.= 574(S)S= A= )*8= (74.8*2*398= (43971S8= *9= /&.= *25U(-S= 1&= 5411.3.8&9.43= )*= +1*:78= 54:7= +47(*7= 1&:94,&2.*_== = $*574):(9.43= .7&.9*2*39= B7:.98= 39*757S9&9.43= :94,&2.*== B1*:78=84:8=8&(= :(:3== &8=&:94,&2*= :94,&2.*==B1*:78=5411.3.8S*8= B7:.98=(((= &8=)*=(429=A=1&:94,&2.*`=)43(= 114,&2.*= 2&3:*11*2*39= B7:.98=(((= &:94,&2*= = &'1*&:=3_= !=5S7.*3(*=)*=(74.8*2*398=(43971S8=(-*?= _=)&2&843.:2 = (+ =B.,_=-*=*9= &79.(1*= == = 1*99*=*=5S7.*3(*=24397*=6:*=1&:94,&2.*=3*89=5&8=5488.'1*=&1478=6:*=1&'8*3(*=)*=).++S7*3(*=*397*= 1&114,&2.*=*9=1&:94,&2.*=+47(S*=24397*=6:.1=3>=&=5&8=)*=(429=&884(.S=A=1&:94,&2.*`=*9=6:*=(*99*= *85T(*= 5*:9= U97*= &:94,&2*_= :(:3= 5411.3.8&9*:7= 3&= S9S= 4'8*7;S= 2&.8= 1*8= +1*:78= 439= S9S= ;.8.9S*8`= 5&7+4.8= *3= 5&79.*8= 2&3,S*8`= *9= )*8= +4:72.8`= )*8= &(&7.*38`= )*8=.42.8*8`=*9=)*8=-&33*9438=439=S9S=4'8*7;S8_='&'8*3(*= )*= 5411.3.*8= 8:7= (*8= .38*(9*8= 5*:9= 8*=51.6:*7= 5&7= 1*:7= 9*=9:7*=a=*11*8=8439=54:)7*:8*8`=*9= 3*=8*=)S9&(-*39=5&8=):= ,>3489T2*_= !3= (438S6:*3(*= 1*8= 5411.3.8&9*:78= 8439= ).++.(.1*2*39= .)*39.+.&'1*8_= 5&:97*= 5&79`= 14786:*= 1*8= +1*:78= 8439=4:;*79*8=)*5:.8=-=/4:78`=1*8=5411.3.*8=5*3)*39`=(*=6:.= 5*:9=+&(.1.9*7=1&:94,&2.*=2&.8=3S(S88.9*=94:9=)*=2U2*=:3= ;.8.9*:7_=!3=(43(1:8.43`=1_=)&2&843.:2=*89=&:94,&2*`=8.1= 3>= &= 5&8= *:= )*= 5411.3.8&9.43= 5*3)&39= 1*8= 57*2.*78= /4:78= )4:;*79:7*=)*=1&=+1*:7=*9=5*:9=&:88.=U97*=&114,&2*_== = B.,:7*=,0_= 411.3.*8=5*3)&39*8=):3*=+1*:7=4:;*79*=)*= _=)&2&843.:2 `=5-494=#_=$4>_==

67 '*8=&1'.348=&55&7&.88*39=(-*?=1&=51:5&79=)*8=51&39*8=84:8=1*++*9=)*=2:9&9.43=543(9:*11*8= &;*(=:3*=+7S6:*3(*=+&.'1*`=)*=147)7*=)*=+=B*1)2&33`=+33+_=1=*89=)43(=574'&'1*=6:*=1*= )S9*72.3.82*= )*= 1&1'.3.82*= )*8= 3*499.S*8= 5:.88*= U97*= ,S3S9.6:*= *9= 6:*= (*8= .3).;.):8= 8:7;.;*39=,7B(*=A=1*:7=2.=49745-.*_=5&38=1&=2*8:7*=41=(*8=.3).;.):8=&1'.348=574):.8*39= )*8= +7:.98`= 1&= 6:*89.43= *89= )*= 8&;4.7= 8.18= 8*= 7*574):.8*39= *++*(9.;*2*39`= *9= 8.18= 8*= 7*574):.8*39=&;*(=1*8=.3).;.):8=;*798=4:=343_=.=1*=)S9*72.3.82*=*89=*3;.7433*2*39&1`=1*8= (43(1:8.438= 5*72*997439= 3S&324.38= )S;&1:*7= 1&= ;&1*:7= 8S1*(9.;*= )*8= &1'.348= *9= )*= (425&7*7=1*:78=(&7&(9S7.89.6:*8=&:==2>(4-S9S749745-*8`=54:7=(4257*3)7*=(*=6:.=5*:9=1*8= 8S1*(9.433*7=4:=(4397*88S1*(9.433*7_== = 841*2*39=7*574):(9*:7=4:=(7>594541>2475-.82*== '*8= &1'.348= 5*:;*398.18= 8*= 7*574):.7*= *9= 8*= 7*574):.8*398.18= &;*(= 1*8= .3);.):8= ;*798== 1-*?= _= )&2&843.:2 `= 1*8= )*:== 9>5*8= ).3).;.):8= +1*:7.88*39= *3= 2U2*= 9*258= *9= 5*:;*39=S(-&3,*7=):=5411*3=B.,_=,-_='*8=(74.8*2*398=(43971S8=*397*=1*8=.3).;.):8=;*798=*9= &1'.348=439=574):.9=)*8=+7:.98=6:*1=6:*=84.9=1*=8*38=):=(74.8*2*39= (+_ &79.(1*= _=!3+.3`=1*8= .3).;.):8= &1'.348= *9= ;*798= 8439= 94:/4:78= 2S1&3,S8= 85&9.&1*2*39= &:= 8*.3= )*8= 545:1&9.438= B.,_= ,-= *9= 5&79&,*39= 1*8= 2U2*8= (-&25.,3438= *1488*= *9= &1_` = ,**.'=b= <:14:= *9= &1_` = ,**/=b= '&).*= *9= &1_` = ,**0_= '&= 897:(9:7*= ,S3S9.6:*= )&57T8= 1&3&1>8*= )*= 574+.18= B'= )*= 545:1&9.438= )*= _= )&2&843.:2 = *9= _= 143,.+41.& = (+ _= &79.(1*= = 24397*= 6:*= 1*8= )*:== 9>5*8= ).3).;.):8= 3*= +472*39= 5&8= )*= 1.,3S*8= .3)S5*3)&39*8`= &:= 3.;*&:= )*= 1*85T(*= (422*= A= 1S(-*11*=)*=1&=545:1&9.43=B.,_=,/_=5*=51:8`=1&=).;*78.9S=,S3S9.6:*=2*8:7S*=)&57T8=)*8= 2&76:*:78= B'`= )43(= )*8= 2&76:*:78= 3*:97*8= *89= 1&= 2U2*= 54:7= 1*8= )*:== ,74:5*8= ).3).;.):8`=(*=6:.=3*=8:,,T7*=5&8=)*=,4:149=)S97&3,1*2*39=8S1*(9.+=51:8=2&76:S=6:*=(-*?= 1*8=;*798_=.=1&1'.3.82*=3S9&.9=5&8=(4397*88S1*(9.433S`=.1=8*7&.9=5&7+4.8=+.=S`=(422*=(-*?=1*8= *85T(*8=2>(4-S9S749745-*8`=47`=(-*?=1*8=*85T(*8=2.=49745-*8`=1&1'.3.82*=*89=543(9:*1=(*= 6:.= 84:1.,3*= 6:.1= *89= 51:8= 574'&'1*2*39= 51:8= 4:= 24.38= (4397*= 8S1*(9.433S_= '&= ).;*78.9S= ,S3S9.6:*= S6:.;&1*39*= )*8= .3).;.):8= ;*798= *9= &1'.348= *9= 1&'8*3(*= )*= ).++S7*39.&9.43= ,S3S9.6:*= *397*= 1*8= )*:== ,74:5*8= ).3).;.):8= 8:,,T7*39= 6:*= 1&1'.3.82*= &55&7&9= 543(9:*11*2*39= )*=+&O43= 7S(:77*39*= )&38= 1*8=*85T(*8= 2.=49745-*8`= (422*= (-*?= )&:97*8= ;S,S9&:=_= 8S&324.38`= )&38= (*79&.3*8= 545:1&9.438= )*= _= )&2&843.:2 `= 1*8= .3).;.):8= &1'.348= 7*57S8*39*39=0*=)*8=.3).;.):8= (+ _= &79.(1*= _=#U2*=14786:*=1*8=&1'.348=)&:97*8=*85T(*8= 8439= (:19.;S8= 8:7= )*8= 2.1.*:== 7.(-*8= *3= 8:(7*8= 6:.= 1*:7= 5*72*99*39= )*= 8:7;.;7*`= .18= 3*= 7*57S8*39*39= 6:*= += )*= 1&= 545:1&9.43= B*1)2&33`= +33+_= 1*99*= ).++S7*3(*= )*= +7S6:*3(*= 5*:9= U97*= &997.':S*= A= 1&= 7*574):(9.43= )*8= .3).;.):8= &1'.348= *9= A= 1&= ).85*78.43= )*= (*= (&7&(9T7*`=51:8=6:A=:3=9&:==)*=2:9&9.43=51:8=S1*;S_='*8=.3).;.):8=&1'.348=574):.8*39=*3= *++*9= )*8= +7:.98`= 6:.= (439.*33*39= )*8= ,7&.3*8`= )439= 1&= 24.9.S= (439.*39= :3= *2'7>43= 8439= +*79.1*8= *9= )439= +*= ,*72*39= &57T8= +2= 24.8= 84:8= 9*77*= (+_ &79.(1*= _= &38= )S24397*7= 1-S7.9&'.1.9S=):=(&7&(9T7*`=(*8=)433S*8=24397*39=6:*=1*8=&1'.348=5*:;*39=8*=7*574):.7*_=5*= 51:8`=1&'8*3(*=)*=).++S7*39.&9.43=,S3S9.6:*=*397*=1*8=.3).;.):8=;*798=*9=&1'.348=8:,,T7*39= 6:.1=3>=&=5&8=)&55&7.*2*39=57S+S7*39.*1=8*143=1*=9>5*=).3).;.):=B.,_=,/_== !3=(43(1:8.43`=1&55&7.9.43=)*8=&1'.348=5*:98U97*=):*=A=1&=+4.8=A=)*8=2:9&9.438=*9=A= 1&=).85*78.43=)*=(*8=2:9&9.438=5&7=1&=7*574):(9.43=B.,_=,.`=24)T1*=-_=1=3>=&=)43(=5&8= ).841*2*39= 7*574):(9*:7= )*8= &1'.348= *9= 1&1'.3.82*= 8*7&.9= )43(= 2&.39*3:= A= 1S9&9= )*= (7>594541>2475-.82*= )&38= 1*8= 545:1&9.438= B.,_= ,._= '&:94,&2.*= 5488.'1*= )*= _= )&2&843.:2 =*3(&)7S=0=5*:9=7S;S1*7=51:8=+7S6:*22*39=)*8=&11T1*8=7S(S88.+8=*9=*=51.6:*7=1&= 51:8=,7&3)*=+7S6:*3(*=)*8=.3).;.):8=&1'.348=(-*?=(*99*=*85T(* =.&'1*&:=0_==

68 = = = = &= '=

= = B.,:7*=,1_= 45:1&9.43= )*=*5-&1&39-*7&=)&2&843.:2 =A=4.,3*;.11*=&`=)&57T8=<:14:= *9=&1_` =,**/`=5-494= _=*7349=*9=A=#439+*77.*7='`=5-494=#_=$4>_=&347&2&=):=8.9*=)*=#439+*77.*7=(`=5-494=#_=$4>_=

69 ._- '&=8:7;.*=)*8=.3).;.):8=&1'.348= = !3=&)2*99&39=6:*=1&1'.3.82*=&55&7&.88*=A=)*=342'7*:8*8=7*57.8*8=*9=84.9=97&382.8`= 1*=+&.9=6:.1=3*=84.9=5&8=+.=S=(-*?=(*8=*85T(*8=8:,,T7*=6:*=(*9=&1'.3.82*=84.9=574'&'1*2*39= (4397*88S1*(9.433S_= &7= (4397*`= .1= &= 5:= U97*= 8S1*(9.433S= (-*?= )*8= *85T(*8= 2>(4-S9S749745-*8`=6:.=54:88*39=*3=84:88'4.8=-:2.)*8=*9=842'7*8_=+.3=)*=(4257*3)7*= 1S;41:9.43= )*= 1&= 2>(4-S9S749745-.*`= .1=+&:9= (425&7*7= 1&= ;&1*:7= 8S1*(9.;*= )*8= )*:== 9>5*8= ).3).;.):8=)&38=)*8=*3;.7433*2*398=9>5.6:*8=)*8=*85T(*8=2>(4-S9S749745-*8=4:=343`=*9= (425&7*7= 1*8= 97&.98= )*8= .3).;.):8= &1'.348= &:== .3).;.):8= ;*798= *9= &:== *85T(*8= 2>(4-S9S749745-*8_= <&.= S9:).S= 1&= 8:7;.*= )*8= &1'.348= )&38= :3= 84:88'4.8= &88*?= 842'7*= *3= !88433*= 4.,3*;.11*`= .&'1*&:= 2= *9= )&38= :3*= 5*:51*7&.*= (:19.;S*= )&38= 1 S7&:19= #439+*77.*7`=.&'1*&:=2=)*=,**0=A=,**2_= = = .9*8= ;4.,3*;.11*=; = 439+*77.*7= = 48.9.43=,S4,7&5-.6:*= .2J,+8`=,J,+!= .-J=-38`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`= 4.,3*;.11*= *9= #439+*77.*7= *9= *++*(9.+= )*8= 545:1&9.438=)*= _=)&2&843.:2 _=$=a=I:&39.9S=)*=7&).&9.43=5-4948>39-S9.6:*=&(9.;*_= = <:14:= *9= &1_ = ,**/= &;&.*39= )S/A= S9:).S= 1&= 545:1&9.43= )*= 4.,3*;.11*= *9= 439= 24397S= 6:*=1*8=)*:==5-S349>5*8=5&79&,*&.*39=:3=(479T,*=2>(47-.?.*3=97T8=).;*78.+.S=,+=*85T(*8= )*= (-&25.,3438= .)*39.+.S*8_= 5&57T8= (*99*= 2U2*= S9:)*`= 1&= 2475-414,.*= )*= 1&55&7*.1= ;S,S9&9.+=&S7.*3=)*8=)*:==5-S349>5*8=3*=).++T7*=5&8=8.,3.+.(&9.;*2*39`=&.38.=6:*=1&=;.9*88*= )*=(74.88&3(*=)*8=9.,*8=b=1*8=.3).;.):8=;*798=7*O4.;*39=./==)*=1*:7=2&9.T7*=47,&3.6:*=)*8= (-&25.,3438_= *1488*= *9= &1_ = ,**.'= &;&.*39= 5&7= &.11*:78= S9:).S= 1*8= 7&(.3*8= )5.5&(9.8= 2.(745->11&` = ;*798= *9= &1'.348`= *9= 24397S= 6:*= 1*8= 8*(43)8= S9&.*39= 51:8= 2>(47-.?S8= 2&.8= 5&79&,*&.*39=1*8=2U2*8=(-&25.,3438=6:*=1*8=.3).;.):8=;*798_=!3+.3`='&).*= *9=&1_ =,**0= 439= 24397S= 6:*= )*8= _= 143,.+41.& = &1'.348= *9= ;*798= S9&.*39= &884(.S8= &:== 2U2*8= *85T(*8= )*= (-&25.,3438`=94:9=(422*=(-*?= 5.5&(9.8=-*11*'47.3* =&12.&`=+323&_=5&57T8=(*8=S9:)*8`=1*8= .3).;.):8= &1'.348= *9= 1*8= .3).;.):8= ;*798= 8439= .3).++S7*3(.&'1*8= *3= 9*72*8= )*= 5&79*3&.7*8= +43,.6:*8=b=/&.=)43(=+4(&1.8S=2&=9-T8*=8:7=1S9:)*=)*8=5&79.*8=&S7.*33*8=)*= _=)&2&843.:2 = ;*798= *9= &1'.348= (+_ &79.(1*= `= 349&22*39= 8:7= 1&= 8:7;.*`= 1*= 8:((T8= 7*574):(9*:7= *9= 1&= 5->8.414,.*=)*8=S(-&3,*8=,&?*:=_=

70 = 42'7*=).3).;.):8 = -/* &=

-**

,/*

,**

+/*

+**

* 19/03/2007 08/04/2007 28/04/2007 18/05/2007 07/06/2007 27/06/2007

&9* = '= (=

_=)&2&843.:2 = &1'.348= _=)&2&843.:2 =;*79=

)= 43):(9&3(*=8942&9.6:*== 241= ,*_=2 8, _8 8+ = 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 0 500 1000 1500 2000 8, 8+ = = = = 241_ 2 _8 = B.,:7*= ,2_= 842'7*= )*= 9.,*8= ;*79*8= 4:= &1'.348= &:= (4:78= )*= 1&= 8&.843= &_= '*8= 9.,*8= (422*3(*39= A= 8S(-*7=)T8=2&.='=*9=1*8=8>2592*8=8&,,7&;*39=*3=/:.3=8:794:9=54:7=1*8=&1'.348=(_=)=#*8:7*=)*= 1&=(43):(9&3(*=*3=7S5438*=A=:3*=&:,2*39&9.43=)*=$=$&).&9.43=-4948>39-S9.6:*=(9.;*=b=3--= 54:7=(-&6:*=5-S349>5*_=-4948=#_=$4>=

71 425&7&.843=)*=1&55&7*.1=;S,S9&9.+=*9=5->8.414,.*=)*8=S(-&3,*8=,&?*:== = #439+*77.*7`= '&= 2475-414,.*= ,S3S7&1*= *89= 1&= 2U2*`= 8&:+= &:= 3.;*&:= )*= 1&= +*:.11*= 6:.= &55&7&9=7S):.9*=*3= 9&.11*=*9=*3= S5&.88*:7= .&'1*&:= 3= (422*= (*11*= )*8= .3).;.):8= &1'.348= ) _= -*11*'47.3* = &12.&`= +323'_= '*:7= 5->8.414,.*= *89= )&.11*:78= ).++S7*39*=a= 1*8= .3).;.):8= &1'.348=7*85.7*39=24.38=A=14'8(:7.9S`=(*=6:.=5*:9=8*=51.6:*7=5&7=:3*=(43(*397&9.43=51:8= +&.'1*= *3= (&7'43*= )*8= +*:.11*8= F.11.&28= ]= B&77&7`= +33*= b= 29-47`= +33/`= *9= 439= :3*= (43):(9&3(*= )*= )*:== A= (.36= +4.8= 51:8= +&.'1*= B.,_= ,2`= .&'1*&:= 3`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`=)&57T8=1 &79.(1*= _=#B=a=2&9.T7*=+7&(-*`=#=a=2&9.T7*=8T(-*`='=a=8:7+&(*=+41.&.7*=5&7=:3.9S= )*=54.)8`=38=a=9*89=89&9.89.6:*=343=8.,3.+.(&9.+=*9=$=9*89=8.,3.+.(&9.+`=0*_*/_= = = 5&79.7= )*= 2&.`= 1*8= +*:.11*8= )*8= .3).;.):8= &1'.348= 57S8*39*39= )*8= 8.,3*8= )*= )S88T(-*2*39=B.,_=,2`=6:.=8S9*3)*39=7&5.)*2*39=A=94:9*=1&=+*:.11*`=5:.8='.*399=,&,3*39= 94:9*=1&=9.,*_=1*8=8.,3*8=&++*(9*39=&:88.=1*8=.3).;.):8=;*798`=2&.8=*3=24.3)7*=5745479.43_=18= )*;.*33*39= 5&79.(:1.T7*2*39= ;.8.'1*8= A= 5&79.7= )*= 2.= 2&.`= 5:.86:*= )*8= 9.,*8= 8T(-*39= *39.T7*2*39= 54:7= *3+.3= ).85&7&97*= B.,_= ,2`= (+ = &79.(1*= _= '*8= (&:8*8= )*= (*= 5-S342T3*= 3439= 5&8= S9S= S1:(.)S*8= 2&.8= (*8= 8.,3*8= )*= )S88T(-*2*39= &55&7&.88*39= (-*?= )&:97*8= *85T(*8=14786:*=1&=+*:.11*=&=9745=(-&:++S=.1,*7= *9=&1_` =+32._=5*8=2S(&3.82*8=5*72*99*39=A= 1&= +*:.11*= )*= 8&)&59*7= A= )*8= &:,2*39&9.438= )*= 1&= 9*25S7&9:7*= 84:;*39= 1.S*= A= :3*= &:,2*39&9.43= )*= 1&= 1:2.348.9S= A= (4:79= 9*72*= .= *3= &:,2*39&39= 1&= 97&385.7&9.43`= ..= *3= ).88.5&39= )*= 1S3*7,.*= 5&7= 1*= 2S9&'41.82*`= *9= 349&22*39= 1&= 5-4947*85.7&9.43= 4:= ...= *3= 574):.8&39= 51:8= )*= 5.,2*398= =S(7&3== (422*= 1*8= (&749S34)*8= 4:= 1*8= =&39-45->11*8`= 349&22*39=1&=Z*&=&39-.3*=5*22.,8)&28=]=)&28`=+33,=b=29-47`=+33/=b=/702&3=]= 5*22.,8)&28`= +33/_= *143= 1&251.9:)*= )*= 1&= 8:7(-&:++*`= 1&= +*:.11*= 5*:9= )43(= 1&= 8:55479*7`=8:'.7=6:*16:*8=)S,B98=97&):.98=5&7=:3*=3S(748*=*9=7*57*3)7*=1&=5-4948>39-T8*= )&38=1*=7*89*=)*=1&=+*:.11*`=4:=8*=3S(748*7=*39.T7*2*39=29-47`=+33/_=

72 &= '=

= = = = B.,:7*=,3_= !39424+&:3*=57S8*39*=8:7=1*8=9.,*8`=+*:.11*8=*9=+7:.98=)*= _=)&2&843.:2 =&1'.348=a=5:(*7438`= (4((.3*11*8=*9=+4:72.8=&`=(-*3.11*='`=(7.6:*9=(=*9=.-42.8*=)_=:(:3*=)*=(*8=*85T(*8=3*=5479*39= )*=5411.3.*8_=1=8&,.9=51:99=)-*7'.;47*8='=*9=(=4:=)*=57S)&9*:78=&=*9=)_=-4948=#_=$4> _=

73 '*8= &1'.348= 84398.18= 51:8= 8*38.'1*8= A= 1&= 8:7(-&:++*= 6:*= 1*8= ;*798== '*8= 2*8:7*8= 543(9:*11*8=)*=9*25S7&9:7*=)*=+*:.11*8=3*=84:1.,3*39=5&8=)*=).++S7*3(*=8.,3.+.(&9.;*=*9=1*8= .3).;.):8=;*798=8*7&.*39=84:;*39=51:8=(-&:)8_=!3=*++*9`=.18=&'847'*39=51:8=)*=1:2.T7*=6:*= 1*8=.3).;.):8=&1'.348=A=(&:8*=)*=1*:78=5.,2*398_==57.47.`=1*8=.3).;.):8=;*798=8439=(&5&'1*8= )*=7S543)7*=A=(4:79=9*72*=A=:3*=8:7(-&:++*=2&.8=6:*3=*898.1=)*8=.3).;.):8=&1'.348==!3= *++*9=1*:7=(439*3:=*3=*&:=)*8=+*:.11*8=*89=51:8=+&.'1*=&.38.=6:*=1*:7=)*38.9S=8942&9.6:*`=1&= 8:7+&(*= )*= 1*:7= +*:.11*= *9= 1*:7= (43):(9&3(*= .&'1*&:= 3_= = 1S(-*11*= )*= 1&= +*:.11*`= .18= 97&385.7*39='*&:(4:5=24.38=6:*=1*8=.3).;.):8=;*798=B.,_=,2`=(*=6:.=1.2.9*=1*:7=(&5&(.9S=)*= 9-*7247S,:1&9.43_=5*=51:8`=.18=3*=8*2'1*39=5&8=).88.5*7=)*=(-&1*:7 =;.&= 1*=2S9&'41.82*`=3.=1&= 5-4947*85.7&9.43`= (&7= 1*:7= 7*85.7&9.43= 3&:,2*39*= 5&8= &;*(= 1&= 1:2.348.9S= *3= (43).9.438= &79.+.(.*11*8= (+_ &79.(1*= _= 1*5*3)&39`= 1&3&1>8*= )*= 1*:78= 5.,2*398= 3&= 5&8= 24397S= )*= 8:7*=57*88.43= )*= ?*&=&39-.3*= 4:= )&:97*8= (&749S34)*8= (+_ &79.(1*= _= .4:9*+4.8`= 1*8= +*:.11*8=S(-&39.11433S*8=3*=57S8*39&.*39=5&8=)*=8.,3*=)*=8:7(-&:++*_= =#439+*77.*7='*8=.3).;.):8=&1'.348=8T(-*39=)43=51:8=6:*=1*8=.3).;.):8=;*798=(&7= 1*:78= (&5&(.9S8= )*= 9-*7247S,:1&9.43= 8439= 51:8= 7S):.9*8_= !3= 7*;&3(-*`= A= 4.,3*;.11*`= 1*8= +*:.11*8= 3*= 57S8*39*39= 5&8= )*= 8.,3*= )*= 8:7(-&:++*= *9= 3*= 942'*39= 5&8_= '*8= &1'.348= 8439= )&.11*:78=51:8=,7&3)8=6:*=(*:==)*=#439+*77.*7= (+_ &79.(1*= _= = &7=&.11*:78`=1*8=&1'.348=8439=51:8=84:;*39=1*8=;.(9.2*8=)*=57S)&9*:78=6:*=1*8=;*798`=A= #439+*77.*7= (422*= A= 4.,3*;.11*`= *9= )*8= .38*(9*8= 5->945-&,*8= 8439= 51:8= +7S6:*22*39= 4'8*7;S8=8:7=1*8=.3).;.):8=&1'.348=B.,_=,3`=.&'1*&:=+*`= (+ =&79.(1*= _=1*=5-S342T3*=&;&.9= )S/A=S9S=4'8*7;S=(-*?=)&:97*8=&1'.348=(&55&9.((.=]=(&55&9.((.`=+332_=1*99*=-*7'.;47.*= (.'1S*=8:7=1*8=&1'.348=5*:9=8*=51.6:*7=.=5&7=1*:7=(4:1*:7=(4397&89S*`=6:.=5*:9=&99.7*7=51:8`= A=1.389&7=1*8=.3+147*8(*3(*8=->54(-742*8=) 77(-.8=2&8(:1& =5472439= *9=&1_` =8:'2.99*)=..= 5&7=1*:78=+*:.11*8=51:8=9*3)7*8=24.38=S5&.88*8`=24.38=)*=(:9.(:1*=*9=)43(=51:8=&55S9&39*8= A:78&7`= ,**-_= 5*= 51:8`= )*8= 5&9-4,T3*8= 9*18= 6:*= 1*8= 8*:)4(*7(48547& = 85_= .3+*(9*39= 57S+S7*39.*11*2*39=1*8=+*:.11*8= )*8= .3).;.):8= &1'.348_= '*8= &1'.348= )439= 1*= ':),*9= (&7'43S= *89=7S):.9=5&7=7&55479=&:==.3).;.):8=;*798`=439=5*:98U97*=24.38=)*=)S+*38*8`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

74 &= 41:2 *=):=+7:.9=(2 -= 1 0.9 0.8 = = 0.7 = = = = 0.6 0.5 = = 0.4 = = 0.3 = = = = 0.2 0.1 = = 0 = = = =

:94= S.9434= 424= S9S74= 43= ,&2.*= ,&2.*= ,&2.*= ,&2.*= (43971S = '= (= 342'7*=)*=,7&.3*8 = =):=342'7*=)*=,7&.3*8 = +** +.** 3* *** *** 38 = +,** 2* 1* +*** 0* 2** /* 0** .* -* .** ,* ,** +* * * 342'7*=)* =+*79.1.9S =,*72.3&9.43 ,7&.3*8 = = B.,:7*= -* _= B*79.1.9S= (425&7S*= )*8= .3).;.):8= &1'.348= *9= ;*798= )*= _= )&2&843.:2 = A= #439+*77.*7_= !++.(&(.9S=)*8=(74.8*2*398=&79.+.(.*18=&=*3=9*72*=)*=;41:2*=)*=+7:.9=574):.9`=(425&7S=&:==+7:.98= 3&9:7*11*2*39= +472S8= *9= 5749S,S8= 84:8= 8&(`= 342'7*= )*= ,7&.3*8= 574):.9*8= '= *9= 54:7(*39&,*= )*= +*79.1.9S=(`=54:7(*39&,*=)*=,7&.3*8=&;*(=:3=*2'7>43=(425&7S=&:=54:7(*39&,*=)*=,*72.3&9.43=+2= 24.8=51:8=9&7)=(`=*2'7>43=,43+1S=*9=2>(47-.?S_== = :94,&2.* =a= +1*:78= *3+*72S*8= 84:8= 8&(`= 14;:1*= *89= +S(43)S= 5&7= ):= 5411*3= )*= 1&= 2U2*= +1*:7_= *.9434,&2.* =a= +1*:78= 5411.3.8S*8= 2&3:*11*2*39`= 4;:1*= +S(43)S= 5&7= ):= 5411*3= )*= 1&= 2U2*= .3+147*8(*3(*_= 424,&2.* =a= +1*:78= 5411.3.8S*8= 2&3:*11*2*39`= 4;:1*= +S(43)S= 5&7= ):= 5411*3= ):3= .3).;.):=).++S7*39=):=2U2*=9>5*=).3).;.):_= S9S74,&2.* =a=+1*:78=5411.3.8S*8=2&3:*11*2*39`=4;:1*= +S(43)S=5&7=):=5411*3=):3=.3).;.):=).++S7*39=):3=9>5*=).3).;.):=).++S7*39_= 43=(43971S =a=+1*:78= 5411.3.8S*8= 3&9:7*11*2*39= *3+*72S*8= 84:8= 8&(= 54:7= S;.9*7= 94:9*= 57S)&9.43_= '*++.(&(.9S= )*8= (74.8*2*398=*89=2*8:7S*=5&7=1*=;41:2*=)*8=+7:.98`=(&7=&:=242*39=)*=1&=7S(419*=:3*=5&79.*=)*8=+7:.98= &;&.9=).85*78S=1*:78=,7&.3*8_=%7=1*8=+7:.98=8T(-*39=1478=)*=1&=).85*78.43=*9=,&7)*39=1*:7=;41:2*=.3.9.&1_= '*=;41:2*=*89=)43(=:3=2*.11*:7=*89.2&9*:7=6:*=1*=54.)8=*3=+.3=)*=8&.843_=

75 425&7&.843=)*=1*++.(&(.9S=)*=1&=7*574):(9.43 = '*8=.3).;.):8=&1'.348`=(422*=1*8=.3).;.):8=;*798=+1*:7.88*39=;*78=1&=2.82&.`=*9=1*8= )*:== 5-S349>5*8= 57S8*39*39= 1*= 2U2*= 342'7*= )*= +1*:78`= 6:.= 8439= 5411.3.8S*8= )&38= 1*8= 2U2*8= 5745479.438= (+_= &79.(1*= _= 1*5*3)&39`= 43= 349*= (-*?= 1*8= .3).;.):8= &1'.348= )*= #439+*77.*7=:3*=51:8=,7&3)*=5745479.43=)*=+7:.98=6:.=&;479*39`=5:.8=942'*39=)*=1&=9.,*_=1*= 5-S342T3*= 8*2'1*= U97*= (477S1S= A= 1&= 57S8*3(*= )*= .42.8*8= 4:= )&(&7.*38= 6:.= 8.389&11*39= )&38= 1*8= +1*:78= *9= *25U(-*39= 5*:98U97*= 1&= 5411.3.8&9.43`= *9= 8:794:9= A= 1.389&11&9.43= )*= 5:(*7438=6:.=8*=+.=*39=8:7=1*8=+7:.98=B.,_=,3_='*8=+7:.98=)*8=.3).;.):8=&1'.348=8439=,=+4.8= 51:8= 7.(-*8= *3= &?49*= 6:*= (*:== )*8= ;*798= (+_= &79.(1*= = (*= 5*:9= 5*72*997*= )&:,2*39*7= 1&55*9&3(*=54:7=1*8=5:(*7438=#&99843`=+32*b= &3,*=]=F*89`=+33._= := (4397&.7*= A= 4.,3*;.11*= 1*8= +7:.98= &1'.348= 8439= 574):.98= )&38= 1*8= 2U2*8= 5745479.438=6:*=(-*?=1*8=.3).;.):8=;*798= (+_ &79.(1*= _=:(:3=5:(*743=3&=S9S=4'8*7;S=8:7= (*99*=545:1&9.43_= = !3+.3`= 1*8= +7:.98= &1'.348= 6:.= 5&7;.*33*39= A= 2&9:7.9S= A= #439+*77.*7 = 574):.8*39= )*8= ,7&.3*8`=2&.8=*3=6:&39.9S=6:&97*=+4.8=24.3)7*=6:*=1*8=.3).;.):8=;*798`=&;*(=:3=54:7(*39&,*= )*2'7>438=;.&'1*8=8.2.1&.7*_=*:1*8=,*=)*8=,7&.3*8=)&1'.348=*39*77S*8=)&38=1*=841`=5:.8= 4'8*7;S*8=)*:==&38=51:8=9&7)`=439=57S8*39S=)*8=8.,3*8=)*=,*72.3&9.43`=(4397*=2*=(-*?=1*8= .3).;.):8= ;*798= B.,_= ,3_= = 4.,3*;.11*`= (*8= 2*8:7*8= 3439= 5&8= S9S= 7S&1.8S*8= *9= 8*7&.*39= 3S(S88&.7*=54:7=(425&7*7=1&=;&1*:7=8S1*(9.;*=)*8=&1'.348=)&38=1*8=)*:==545:1&9.438_= = :='.1&3`=(*8=2*8:7*8=5*72*99*39=)S9&'1.7=:3*=(425&7&.843=)*=1&=;&1*:7=8S1*(9.;*= +*2*11*=)*8=.3).;.):8=&1'.348=5&7=7&55479=&:==;*798=)&38=1*8=)*:==545:1&9.438=.&'1*&:=++`= B.,_=,3_=1=7*88479=6:*=1&=;&1*:7=8S1*(9.;*=+*2*11*=)*8=.3).;.):8=&1'.348=*89='*&:(4:5=51:8= +&.'1*= 6:*= (*11*= )*8= .3).;.):8= ;*798= A= #439+*77.*7`= 9&3).8= 6:*11*= *89= S,&1*= A= (*11*= )*8= .3).;.):8=;*798=A=4.,3*;.11*_= = = 439+*77.*7= = ;4.,3*;.11*= :7;.*=,14'&1*= C== -==== = = :7;.*=)*=1&=9.,*=&:=242*39=)*=1&=+7:(9.+.(&9.43$= C= ,== 1= = = C= -= = 842'7*=)*=+1*:78= C== -=== C= -= = 842'7*=)*=+7:.98= C== -=== C= -= = .&.11*=)*8=+7:.98= C=,== 1== = C= -= = 842'7*=)*=,7&.3*8= C=.== 1=== = = = C.&'.1.9S=)*8=,7&.3*8= C= .== 1= = = = = = &1*:7=8S1*(9.;*= C= 0.== 1= = = C= -= = = &'1*&:=+-_= 1425&7&.843=)*=1&=;&1*:7=8S1*(9.;*=+*2*11*=)*8=.3).;.):8=&1'.348=*9=)*8=.3).;.):8=;*798= A= #439+*77.*7= *9= 4.,3*;.11*_= $1422*= 1&= 24.9.S= )*8= 9.,*8= ).85&7&9= *3= /:.3= 54:7= 1*8= &1'.348`= 1*:7= 8:7;.*=*3=/:.3=*89=)*:==+4.8=51:8=+&.'1*`= (+ _= &79.(1*= _= = =

76 &=

842'7*=). 3).;.):8 = ,**0 = ,**1 = ,**2 = /** .**

-** ..3 ,** .,-

+** 33 1- * -+ -+ -24 -15 -318 -70 8+** 8,** ,=) = +*=) = 8-** 8.**

'= 842'7*=). 3).;.):8 =

13 / 10 10 10 9 6

* /=) = -3 -5 -5 8/ -1 -12

8+* _=)&2&843.:2 = ;*79 = 8+/ 57S8*39=1&33S*=3= &1'.34 = =

;*79 = &'8*39=5&7=7&55479=A=1&33S*=38+= &1'.34 = = )= 3).;.):=)472&39=4:=A=+147&.843=84:9*77&.3*=1&33S*=3= ;:=*3=38+=*9=3(+=

= B.,:7*=-+_= 5>3&2.6:*=)*8=)*:==545:1&9.438=)*= _=)&2&843.:2 `=A=#439+*77.*7=&=*9=A=4.,3*;.11*='= *397*= ,**0= *9= ,**2_= '*8= .3).;.):8= (4259S8= 548.9.;*2*39= 8439= ;:8= 9&3).8= 6:*= (*:== (4259S8= 3S,&9.;*2*39=8439=&'8*398_='*8=.3).;.):8=)*=,**0`=343=;.8.'1*8=*3=,**1=2&.8=7S&55&7:8=*3=,**2=8439= (438.)S7S8= (422*= )472&398= 5&8= )*= 9.,*= &33:*11*`= 4:= 51:8= 7&7*2*39= 574):(9.43= ):3*= 9.,*= 84:9*77&.3*_ = =

77 ._. '*=2&.39.*3=)*=1&1'.3.82*= = '&1'.3.82*= *89= 2&.39*3:= )&38= 1*8= *85T(*8= 2.=49745-*8$= 5&7= 1&= 2:9&9.43= *9= 1&= ).85*78.43=):=(&7&(9T7*`=A=1S9&9=)*=(7>594541>2475-.82*_='&=).85*78.43=)*=(*=(&7&(9T7*`=*9= )43(=8&=7*574):(9.43`=*89=)S5*3)&39*=)*=1&=8S1*(9.43=*9=)*=1&=)S7.;*_=%7`=1&=+&.'1*=8:7;.*= )*8=&1'.348=84:1.,3*=1*:7=(4397*88S1*(9.43`=&:=24.38=A=#439+*77.*7_= 1*5*3)&39`=(*79&.38=+&(9*:78=8*2'1*=9*2547.8*7=(*99*=(4397*88S1*(9.43=*9=5*72*997*= )*=2&.39*3.7=(*9=&1'.3.82*_=.4:9=)&'47)`=&:=3.;*&:=)*=1.3).;.):`=8.1=*89=;7&.=6:*=1&=9.,*= )*=1&33S*=).85&7&9=7&5.)*2*39`=(*1&=3.251.6:*=5&8=1&=2479=)*=1.3).;.):_=!3=*++*9`=94:9= (422*=1*8=.3).;.):8=;*798`=.1=8&,.9=)*=51&39*8=5S7*33*8`=*9=1*8=.3).;.):8=5*:;*39=574):.7*= :3*=34:;*11*=9.,*=1*8=&33S*8=8:.;&39*8= (+_ &79.(1*= =b=-*++*7843= *9=&1 _`=,**/& =b= '&).*= *9=&1 _`= ,**0_=5&38=)&:97*8=*85T(*8`=1*8=.3).;.):8=&1'.348=8439=84:;*39=;:8=51:8.*:78=&33S*8=)*= 8:.9*='&).*= *9=&1_` =,**0=b=&12.&`=+323'_=&7=&.11*:78`=1*8=.3).;.):8=&1'.348=8*2'1*39=U97*= 51:8=342'7*:==A=5&88*7=:3*=&33S*=84:8=9*77*=6:*=1*8=.3).;.):8=;*798=B.,_=-+`= (+ =&79.(1*= _= '*= 5&88&,*= ):3*= &33S*= 84:9*77&.3*`= 84.9= 5&7(*= 6:*= 1.3).;.):= *89= )472&39= 4:= 24.38= 574'&'1*2*39=5&7(*=6:*=1&=+147&.843=*89=84:9*77&.3*`=*89=:3=5-S342T3*=+7S6:*39=(-*?=1*8= 47(-.)S*8`= *9= 349&22*39= 1*8= 2>(4-S9S749745-*8`= (422*= 5.54,.:2= &5->11:2 = :22*7-&>*8`= +3/+=b= 4>7.30.`= +321=b= $4'.3`= +323_= 1-*?= .89*7&= 4;&9& `= 1&= )472&3(*= *89= +7S6:*22*39=4'8*7;S*=1&33S*=8:.;&39=1&=574):(9.43=):3*=-&25*=+147&1*`=.1=8&,.7&.9=):3= 24>*3=)S;.9*7=1S5:.8*2*39=)*8=7*884:7(*8=)*=1&=51&39*`=:3*=&33S*=8&''&9.6:*=84:9*77&.3*= *3=6:*16:*=8479*=7?4804`=,**,=b=-*++*7843`=,**/&_=1*99*=)472&3(*=&=3S&324.38=:3=(429`= 94:9= ) &'47)= 5&7(*= 6:*11*= 8:557.2*= 1&= 7*574):(9.43= 8*=:S*= 5*3)&39= :3= &3= *9= *38:.9*= 5&7(*=6:*=1&=2479&1.9S=8*2'1*=51:8=S1*;S*=5*3)&39=1*8=5-&8*8=84:9*77&.3*8=-*++*7843= *9= &1_` =,**/&=b=,**0=b=-*++*7843`=,**0_= = :=3.;*&:=)*=1*85T(*`=8.=1*=(&7&(9T7*=*89=97&382.88.'1*`=.1=1*89=5*:=5&7=1*8=,7&.3*8`= 6:.= 8439= 574):.9*8= *3= 24.3)7*= 6:&39.9S= *9= 8439= 24.38= ;.&'1*8_= 1*5*3)&39`= 1*= 8S(-&,*= (422*=1-*7'.;47.*=94:(-*39=1*8=+*:.11*8=57.3(.5&1*2*39=*3=/:.3`=*9=1&=+147&.843=&=1.*:=2.8 2&.`=1S,T7*2*39=51:8=57S(4(*2*39=(-*?=1*8=.3).;.):8=&1'.348= (+_ &79.(1*= _='*=5411*3=)*8= .3).;.):8= &1'.348= 5*:9= )43(= U97*= ).85*78S= A= (*= 242*39`= *9= 5*72*997*= )*= 5411.3.8*7= )*8= .3).;.):8= ;*798`= 6:.= 5479*7439= 1*8= +7:.98= *9= ).85*78*7439= 1*= (&7&(9T7*= &1'.348_= '*8= (74.8*2*398=(43971S8=8439=;.&'1*8=B.,_=-*`=*9=1*8=+7:.98=;*798=3*=8439=5&8=8.,3.+.(&9.;*2*39= 51:8= 5*9.98= 8.18= 8439= 5411.3.8S8= 5&7= ):= 5411*3= &1'.348= 6:*= ):= 5411*3= ;*79`= (*= 6:.= 1&.88*= 8:5548*7=:3*=6:&39.9S=*9=:3*=6:&1.9S=S,&1*=)*=5411*3_='*=5411*3=)*8= _=-*11*'47.3* =&1'.348= *89=)&.11*:78=.)*39.6:*=*3=9&.11*=A=(*1:.=)*8=;*798=&12.&`=+323'_='&=6:&39.9S=)*=(42548S8= ;41&9.18= 57S1*;S8= *89= +&.'1*= *9= 343= ).++S7*39.&'1*= *397*= 5-S349>5*_= 5*8= (42548S8= 2.347.9&.7*8=8439=94:9*+4.8=574):.98=:3.6:*2*39=(-*?=1*8=.3).;.):8=;*798`=(*=6:.=54:77&.9= 1*8= ).++S7*3(.*7`= *9= *=51.6:*7&.9= 1&= 51:8= ,7&3)*= 5745479.43= )*= +1*:78= 5411.3.8S*8= (-*?= (*8= .3).;.):8= (+_ &79.(1*= _= 1*99*= ).85*78.43= ):= 5411*3= 24397*=6:*= 1&= ;&1*:7= 8S1*(9.;*= 2B1*= )*8=&1'.348=*89='.*3=8:5S7.*:7*=A=1&=;&1*:7=8S1*(9.;*=+*2*11*=(&1(:1S*=57S(S)*22*39_= '&= 897:(9:7*= 85&9.&1*= )*= 1&= 545:1&9.43= 3*= 57S8*39*= 5&8= )*= 8.,3*8= )&,7S,&9.43= 57S+S7*39.*11*= )*8= .3).;.):8= &1'.348= 4:= )*8= ;*798= *9= 1*8= .3).;.):8= ;*798= (422*= 1*8= .3).;.):8= &1'.348= 24397*39= 6:*= 1&= 51:5&79= )*8= .3).;.):8= 8439= ).89&398= )*= +*8+/= 2= (+_ &79.(1*= _= 1*99*= 897:(9:7*= 85&9.&1*= 3*89= 5&8= (477S1S*= A= 1&= 897:(9:7*= ,S3S9.6:*`= (*= 6:.= )S24397*= 6:*= 1*8= )*:== 5-S349>5*8= ).85*78*39`= )*= 1&= 2U2*= +&O43`= 1*:78= ,7&.3*8= 4:= 1*:7= 5411*3=A=143,:*=).89&3(*_=

78 =

3).;.):8=;*798= 3).;.):8=&1'.348== 2U2*=2>(47-.?&9.43= 2U2*=2>(47-.?&9.43= 5-4948>39-T8*= 5&8=)*=5-4948>39-T8*=

':),*9=1&7'43S=1= = ':),*9 =1 = 5.,2*398 = )*38.9S=8942&9.6:*= 574):(9.43=)*=9.,*= (:9.(:1*= 2S9&'41.82*=7S):.9=

(43):(9&3(*= 9&.11*=)*=1&=9.,* = 8942&9.6:*= =

4) *:78 = 342'7*=)*= +1*:78 = 5S+*38*=(4397*= 5S+*38*=(4397*=1&= 5&9-4,T3*8= 8:7(-&:++* = *9= -*7'.;47*8= 5411.3 .8&9.43 = 342'7*=)*=+7:.98= 5488.'1*8 = 8:7;.*=)*=1&=9.,*=A=1&= +7:(9.+.(&9.43=

5411.3.8&9.43=)*8= 5411.3.8&9.43=)*8=+1*:78=)*8=.3).;.):8=&1'.3 48 = +1*:78=)*8=.3).;.):8= ;*798 = =,7&.3*8 = 574):(9.43=)*=,7&.3*8 = 8:7;.*= 84:9*77&.3*=

S,*3)* =a= B&(9*:78= 2479&1.9S= ,*72.3&9.43 = )472&3(*= ,*72.3&9.43 = *=97.38T6:*8= 84:9*77&.3*= = <=a =5.2.3:9.43=)*= <=(-*?=1*8==== &1'.348=5&7=7&55479=&:== 4:;*&:== 2U2*= 4:;*&:== .3).;.):8=;*798= .3).;.):8= .3).;.):= .3).;.):8= =<=a= I:&39.9S=)*= <=S,&1*=(-*?= ;*798=)439= &1'.348= &1'.348=4:= 1*8=&1'.348=A=(*11*=)*8= (*79&.38= ;*798=8.= .3).;.):8=;*798=4:=8:5S7.*:7*= 5479*:78=)*= -S9S74,&2.* = 3).;.):8=&1'.348= 1&1'.3.82*= 3).;.):8=;*798 = = B.,:7*=-,_= .1&3=)*8=+&(9*:78=&++*(9&39=1&=;&1*:7=8S1*(9.;*=)*8=.3).;.):8=&1'.348=&:=(4:78=)*=1*:7=;.*= (425&7S=&:==.3).;.):8=;*798_='&=)472&3(*=&=:3=*++*9=548.9.+=A=1S(-*11*=)*=1.3).;.):=2&.8=3S,&9.+=A= 1S(-*11*=)*=1*85T(*`=5:.86:*11*=).2.3:*=1&=574'&'.1.9S=)*=8*=7*574):.7*_=

79 !3= (43(1:8.43`= 1&= )472&3(*= 5*:9= 5*72*997*= 1&= 8:7;.*= )*8= .3).;.):8= &1'.348= 8:7= 51:8.*:78=&33S*8=*9=1&=).85*78.43=):=5411*3=5*:9=5*72*997*=1*=2&.39.*3=):=(&7&(9T7*=)&38= 1*85T(*_= '*8= &11T1*8= )S1S9T7*8= 8439= )&.11*:78= 51:8= ).++.(.1*8= A= S1.2.3*7= 14786:.18= 8439= ).85*78S8=5&7=1*=5411*3`=5:.86:.18=8439=(&(-S8=(-*?=1*8=-S9S74?>,49*8='45*?= *9=&1_` =,**2=*9= 3*= 8439= 5&8= (4397*88S1*(9.433S8= A= 1S9&9= )*= ,7&.3*_= 1*99*= ).85*78.43= 5&7= 1*= 5411*3= 5*:9= 5*72*997*= A= 1&1'.3.82*= )U97*= &.38.= 2&.39*3:= )&38= 1*8= 545:1&9.438= 84:8= +472*= ):3= 541>2475-.82*= (7>59.6:*`= 7S;S1S= (-*?= 1*8= -424?>,49*8_= '&= 6:&39.9S= )*= 5411*3= 3&= 3S&324.38= 5&8= S9S= 2*8:7S*`= *9= 8*7&.9= 3S(*88&.7*= 54:7= 8&;4.7= 8.= 1.3;*89.88*2*39= )&38= 1&= 7*574):(9.43 =;.&= 1*=5411*3=*89=51:8=.25479&39=(-*?=1*8=.3).;.):8=&1'.348_= = 5:3= &:97*= (9S`= 1&= +&.'1*= ;&1*:7= 8S1*(9.;*= )*8= &1'.348= 5*:9= U97*= (477S1S*= A= 1*:7= ':),*9=(&7'43S=51:8=+&.'1*`=349&22*39=1478=)*=1&=+7:(9.+.(&9.43=41=1*8=.3).;.):8=;*798=439= 51:8= )&554798= )*= 1&= 5-4948>39-T8*_= 1*= ':),*9= (&7'43S= 51:8= +&.'1*= 5*:9= ).2.3:*7= 1&= 574):(9.43= )*= 9.,*= *9= 1*:7= 9&.11*`= &.38.= 6:*= 1*= 342'7*= )*= +1*:78`= *9= 1&= 6:&39.9S= )*= +7:.98= 574):.98= B.,_= -,_= 5*= 51:8= 1*= )*88T(-*2*39= )*8= +*:.11*8= *9= 1-*7'.;47.*= &++*(9*39= &:88.= 1&= ;&1*:7=8S1*(9.;*=)*8=.3).;.):8=&1'.348_=1*8=)*:==)*73.*78=+&(9*:78=3*=8439=5&8=.397.38T6:*8= *9=8439=)S5*3)&398=)*=1*3;.7433*2*39`=1&=(425&7&.843=&;*(=1&=545:1&9.43=)*=4.,3*;.11*= *3= *89= 1&= 57*:;*_= '*= 8.9*= )*= 4.,3*;.11*`= 6:.= 7*88*2'1*= 51:8= &:== +47U98= 41= 54:88*39= )*8= *85T(*8=2>(4-S9S749745-*8`=*89=24.38=)S+&;47&'1*=&:==.3).;.):8=&1'.348`=(*=6:.=8:,,T7*= 6:*=1&=2>(4-S9S749745-.*=3&=5:=U97*=+.=S*=6:*=)&38=)*=9*18=84:88'4.8_= = '&= 51:5&79= )*8= +&(9*:78= S9:).S8= &++*(9*39= 3S,&9.;*2*39= 1&= ;&1*:7= 8S1*(9.;*= )*8= .3).;.):8= &1'.348= 2&.8= 3S1.2.3*39= 5&8= 3S(S88&.7*2*39= 1&1'.3.82*= 54:7= &:9&39_= 5&38= 1&= 2*8:7*= 41= 34:8= &;438= *89.2S= 1&= 51:5&79= )*8= 5&7&2T97*8= 6:.= )S(7.;*39= 1&= 8:7;.*= )*8= .3).;.):8=&1'.348`=9&39=A=1S(-*11*=)*=1.3).;.):=6:*=)*=1&=545:1&9.43`=.1=8*7&.9=2&.39*3&39= 5488.'1*= )*= +472&1.8*7=(*=24)T1*=B.,_= -,`=--`=*3=>=.39S,7&39= 1&=2:9&9.43=6:.= &/4:9*= )*8= .3).;.):8=&1'.348=54:7=57S;4.7=A=143,=9*72*=*9=8*143=).++S7*398=*3;.7433*2*39`=1&=(&5&(.9S= )*8=*85T(*8=2.=49745-*8$=A=+.=*7=1&1'.3.82*_= = = =

B.,_ =-+ = 8S1*(9.43 = 55&7.9.43=*9=2&.39.*3=&:=3.;*&:=)*=1*85T(* =

7*574):(9.43 = + 2:9&9.43 = (7>594541>2475-.82* = 2:9&9.43 =

= B.,:7*=--_= 55&7.9.43=*9=2&.39.*3=)*8=&1'.348=)&38=1*8=*85T(*8=2.=49745-*8`=5&7=7&55479=A=1&=B.,_=,-`= 8*:1=1*=24)T1*=-=&=S9S=7*9*3:=7*574):(9.43=(=2:9&9.43=&.38.=6:*=1->549-T8*=):3=2&.39.*3=84:8= +472*=)*=(7>594541>2475-.82*=)&38=1*85T(*=2.=49745-*_======

80 = = = =)*8=(-&25.,3438=.)*39.+.S8=)&38=1*8=7&(.3*8 = &_=&7.&9.438=.397&85S(.+.6:*8= +** *9=.397&545:1&9.433*11*8=):= 3* (479T,*=)*=2>(47-.?*8= 2* = 1* 0* !=_=1-&25.,3438= *(942>(47-.?.*38=.)*39.+.S8= /* )&38=)*8=7&(.3*8=)*== .* _=)&2&843.:2 =A=4.,3*;.11*= -* ,* +* *

+ + , + , + , 3).;.):8 =

'_=&7.&9.438=.39*785S(.+.6:*8= ):=(479T,*=)*=2>(47-.?*8= = 85T(*8 = = = ) = 1 = 7 = - = ) = & = 2 = &8.).42>(T9* 8= 1479.3&7.&( S*8 ======>2*34,&89*7&( S*8 ======.-S1S5-47&(S*8 ======$:88:1&(S*8 ======*'&(.3&1*8 ======8(42>(T9* 8= ":'*7 = = = = += = = += = 1.1(4=.3& = = = = = += += = = *?.?&1*8 ======+=

= Figure 34. C&7.&9.438= .397&545:1&9.433*11*8= *9= .397&85S(.+.6:*8= ):= (479T,*= )*= 2>(47-.?*8= (-*?= _= )&2&843.:2 =&=*9=;&7.&9.438=.39*785S(.+.6:*8=(-*?=1*8=3*499.S*8='=a= _=)&2&843.:2 =1)`=<:14:= *9=&1_` = ,**/`= _= 143,.+41.& = 11`= '&).*= *9= &1_` = ,**0`= _= 7:'7& = .)&7943)4= *9= &1_` = ,**.`= _= -*11*'47.3* = !-`= .)&7943)4= *9=&1_` =,**.`= _=).89&38 =!)`=.)&7943)4= *9=&1_` =,**.`= _=&9747:'*38 =!&`=.)&7943)4= *9=&1_` = ,**.= *9= _= 2.(745->11& = !2`= *1488*= *9= &1_` = ,**.'_= '*8= (-&25.,3438= 2&/47.9&.7*8= 8439= .3).6:S8= *3= 34.7`=1*8=2.347.9&.7*8=*3=,7.8=5B1*_= =

81 C_ 'S;41:9.43=)*=1&=2>(4-S9S749745-.*=(-*?=1*8= 47(-.)S*8=7*;.8.9S*= = /_+ '*=2&.39.*3=)*8=2.=49745-*8= = '&1'.3.82*= 8*2'1*= 2&.39*3:= (-*?= )*8= *85T(*8= 2.=49745-*8$= A= 1S9&9= )*= (7>594541>2475-.82*`='.*3=6:.1=84.9=(4397*88S1*(9.433S`=2&.8=6:*3=*898.1=)*=1&=8S1*(9.43= )*= 1&= 2.=49745-.*== '&= 2.=49745-.*= &= S9S= )S(4:;*79*= 97T8= 7S(*22*39= *9= 1*= 342'7*= )*85T(*8= 2.=49745-*8$= 3*= (*88*= )&:,2*39*7= (+ _= &79.(1*= _= '*8= 84:88'4.8= 9*25S7S8= 7*,47,*39= )*85T(*8= 2.=49745-*8$= Z.22*7= *9= &1 _`= ,**1= 6:.= 7*O4.;*39= :3*= 5&79= )*= 1*:7= (&7'43*=47,&3.6:*=)*=1*:78=(-&25.,3438=2>(47-.?.*38_='*=(429=)*=(*8=2.=49745-*8$=54:7= 1*8=(-&25.,3438=3&=/&2&.8=S9S=S;&1:S=*9=8*7&.9=3S(S88&.7*=54:7=(4257*3)7*=1*=2&.39.*3=)*= (*8= 2.=49745-*8$=*9= 1*:7= &'43)&3(*= *3= 9*72*8=).3).;.):8=*9= )*85T(*8`= 2&.8= 1*8= (4298= 8439=).++.(.1*8=A=S;&1:*7=54:7=1*8=(-&25.,3438`=):3*=5&79=(&7=.18=8439=).++:8=)&38=1*=841`=*9= .1=)*;.*39=51:8=).++.(.1*=)*=).89.3,:*7=1*8=.3).;.):8`=)*=2*8:7*7=1*:7='.42&88*`=*9=)&:97*= 5&79= 5&7(*= 6:*= 1&'43)&3(*= )*8= (&7545-47*8= *3= 8:7+&(*= 3*= 7*+1T9*= 3.= 1&= '.42&88*= 84:9*77&.3*`=3.=1S9&9=5->8.414,.6:*=):=(-&25.,343=*3(&)7S=/`=$.(-&7)= *9=&1 _`=,**/_= 8S&324.38`= 1*8= 2.=49745-*8$= 8*2'1*39= 24.38= (429*:8*8= 6:*= 1*8= 2>(4-S9S749745-*8= 54:7= 1*:78= (-&25.,3438= 2>(47-.?.*38_= .4:9= )&'47)`= 1*8= 2.=49745-*8$=3*=7*O4.;*39=6::3*=5&79.*=)*=1*:7=(&7'43*=)*8=(-&25.,3438=2>(47-.?.*38`= *9=343=1&=949&1.9S_='*=)*,7S=)-S9S749745-.*=*89=,S3S7&1*2*39=;4.8.3=)*=/*=.&'1*&:=1=*9= ;&7.*=&:=8*.3=):3*=*85T(*=*397*=1*8=8.9*8=*9=&:=(4:78=)*=1&33S*_='*8=51&39*8=2.=49745-*8$= 8439=97T8=-S9S749745-*8=&:=242*39=)*=1&=(74.88&3(*=)*=1&=9.,*=5:.8=)*;.*33*39=)*=51:8=*3= 51:8=&:949745-*_=1*99*=;&7.&9.43=9*2547*11*=)*=1:9.1.8&9.43=)*=(&7'43*=+43,.6:*=5*:9=&:88.= 7S):.7*= 1*= (429= )*= 1&= 2.=49745-.*= 54:7= 1*8= (-&25.,3438_= !38:.9*`= &:= 8*.3= ):3*= *85T(*= (422*= _=)&2&843.:2 `=1*8=.3).;.):8=8439=&884(.S8=A=:3=,7&3)=342'7*=)*=(-&25.,3438`=*9= 94:98=1*8=.3).;.):8=3*=5&79&,*39=5&8=1*8=2U2*8=(-&25.,3438=B.,_=-._=!3=(438S6:*3(*`=1*8= (4298=8439=7S5&79.8=8:7=1*8=).++S7*398=(-&25.,3438=*(942>(47-.?.*38=6:.=1*8=*394:7*39`=(*= 6:.= 5*:9= 5*72*997*= )*= 7S):.7*= 1*= (429= 54:7= (-&6:*= (-&25.,343_= 1*79*8= 1*8= 545:1&9.438= 8439= 84:;*39= )*38*8= (422*= A= #439+*77.*7= *9= )*= 8:7(749`= 51:8.*:78= *85T(*8= )*= 2.=49745-*8$=(4-&'.9*39=+7S6:*22*39=(422*=A=#439+*77.*7=*9=A=4.,3*;.11*`=<:14:= *9=&1_` = ,**/`=2&.8=*3=(425&7&39=1*8=(-&25.,3438=&884(.S8=&:== 5.5&(9.8 =*9=&:== *5-&1&39-*7& `=1&= ).++S7*3(*=*89=+7&55&39*_=.4:9*8=1*8=*85T(*8=+47*89.T7*8=) 5.5&(9.8 =8439=&884(.S*8=343=5&8=A= )*8= &8.).42>(T9*8= *(942>(47-.?.*38= (422*= 1*8= *5-&1&39-*7& `= 2&.8= A= )*8= 8(42>(T9*8= *(942>(47-.?.*38= B.,_= -.`= .)&7943)4= *9= &1_`= ,**.=b= *1488*= *9= &1_` = ,**.'=b= .)&7943)4= ]= $*&)`=,**2_=.*3=6::3*=S9:)*=8:7=:3=2U2*=8.9*=8*7&.9=3S(*88&.7*=54:7=(43(1:7*`=43=5*:9= 3S&324.38= 574548*7= 6:*= )&38= :3= 2U2*= -&'.9&9`= (*8= )*:== ,*37*8= )*= 2.=49745-*8$= 3*= 5&79&,*39= 5&8= 1*8= 2U2*8= 5&79*3&.7*8= 2>(47-.?.*38`= (*= 6:.= ).1:*= 1*8= (4298= )*= 1*:7= 2.=49745-.*_= 1*99*= ).++S7*3(*= *397*= 1*:78= (479T,*8= )*= 2>(47-.?*8= 5*:9= U97*= .39*757S9S*= (422*=7S8:19&39=):3*=*=(1:8.43=(425S9.9.;*=6:.=1*8=+47(*=A=).;*78.+.*7=1*:78=5&79*3&.7*8= F*''`=,**,=*9=2*3&39=A=1&=+472&9.43=):3*=,:.1)*_= '*=2&.39.*3=)*8=2.=49745-*8$=,S3S7&1.89*8=*89=)43(=574'&'1*2*39=5*72.8=)&38=)*8= +47U98= 41= 1*8= -9*8= 8439= 342'7*:== A= *397*9*3.7= )*8= 1#8=b= )*= 1A`= (422*39= 1&= 2>(4-S9S749745-.*=5*:98*11*=S2*7,*7==

82 = = = = =

+=( 2= ,=(2 = *_/ =(2 = = = = Figure 35. 1425&7&.843=)*= _=)&2&843.:2 =&1'.34`=*9=)*=)*:==*85T(*8=2>(4-S9S749745-*8`= _=&:89.3&* = *9= _=*=.,:&= (-*?=6:.=1*8=+*:.11*8=5*78.89*39=2&.8=8439=7S):.9*8=5&7=7&55479=&:==*85T(*8=2.=49745-*8$= )*= *5-&1&39-*7& _=

83 /_, 5*=1&=2.=49745-.*=A=1&=2>(4-S9S749745-.*= = = '*8=.3).;.):8=&1'.348=34:8=24397*39=6:.18=8439=349&22*39=(4397*=8S1*(9.433S8=A= (&:8*=)*=1*:78=+*:.11*8`=6:.=3*=1*:7=7&55479*39=&:(:3='S3S+.(*=*9=6:.=1*8=+7&,.1.8*39=)&38=)*8= *3;.7433*2*398= 8*(8_= %7`= :3= )*8= 97&.98= (422:38= &:== *85T(*8= 2>(4-S9S749745-*8= *89= 1&= 7S):(9.43= )*8= +*:.11*8`= *9= 84:;*39= 1&= 5*79*= )*8= 8942&9*8= '*&0*`= +33._= '&= 7S):(9.43= )*8= +*:.11*8= *89= 51:8= 4:= 24.38= .25479&39*= (-*?= 1*8= *85T(*8= 2>(4-S9S749745-*8`= (-*?= *5-&1&39-*7&=&:89.3&* =*9= _=*=.,:& =*11*8=8439=*3(47*=)S;*1455S*8`=2&.8=.=+4.8=24.38=,7&3)*8= 6:*= (-*?= )*8= *85T(*8= 2.=49745-*8$= 574(-*8= B.,_= -/_= '*8= .3).;.):8= &1'.348= 57S8*39*39= &:88.=)*8=+*:.11*8=7S):.9*8=5&7=7&55479=&:==.3).;.):8=;*798`=2&.8=*11*8=8*2'1*39=*3(47*=9745= (429*:8*8`=&:=24.38=)&38=:3=*3;.7433*2*39=8*(_=5&38=1&=545:1&9.43=)*=4.,3*;.11*`=6:.=&= :3=2.(74(1.2&9='*&:(4:5=51:8=-:2.)*=6:A=#439+*77.*7`=1*8=+*:.11*8=3*=57S8*39*39=5&8=(*8= 8.,3*8=)*=)S88T(-*2*39`=*9=1*8=.3).;.):8=&1'.348=8439=(433:8=8:7=(*99*=545:1&9.43=)*5:.8=/= &38`=41=.18=7*57S8*39*39=*397*=.*=*9=0*=)*8=.3).;.):8_=5*=51:8`=1*8=(&8=)&1'.348=8439=51:8= +7S6:*398= )&38= 1*8= +47U98= )*= 8472&3).*= 4:= )11*2&,3*= 6:*3= 7S,.43= 2S).9*77&3S*33*= .&'1*&:=2_='*8=545:1&9.438=41=1&1'.3.82*=&=5:=U97*=+.=S=A=5&79.7=)*85T(*8=2.=49745-*8$= 8439= 574'&'1*2*39= 8.9:S*8= )&38= )*8= +47U98= 842'7*8= *9= -:2.)*8`= A= 1.389&7= )*= (*11*8= 41= ;.;*39=)*8=*85T(*8=2>(4-S9S749745-*8_== = 5&:97*=5&79`=1*8=*85T(*8=2>(4-S9S749745-*8=8439=7&7*2*39=):3='1&3(=.22&(:1S`=*9= 1&=).;*78.9S=)*=1*:7=(4:1*:7=54:77&.9=(477*8543)7*=A=1&=57S8*3(*=)*=8:'89&3(*8=8*(43)&.7*8= 5749S,*&39=)*=1-*7'.;47.*=4:=)*=1&=1:2.T7*='*&0*=+33._=%7=1-*7'.;47.*=*89=*++*(9.;*2*39= :3= )*8= +&(9*:78= 6:.= &++*(9*= 1*= 51:8= 1*8= .3).;.):8= &1'.348`= A= #439+*77.*7= (422*= A= 4.,3*;.11*_='*8=&1'.348=(422*=1*8=2>(4-S9S749745-*8=8439=)&.11*:78=51:8=7.(-*8=*3=&?49*= 6:*=1*8=.3).;.):8=;*798= (+_ =&79.(1*= `=Z.22*7= *9=&1 _`=,**2=*9=U97*=51:8=&55S9&398=54:7=1*8= -*7'.;47*8_=&7=&.11*:78=1&=1:2.T7*=*=(*88.;*=2T3*=&:=5-S342T3*=)*=8:7(-&:++*=6:.=5*:9= *=51.6:*7= 1*= )S88T(-*2*39= 4'8*7;S= (-*?= 1*8= &1'.348`= *9= 54:7= 8*3= 5479S,*7= 1*8= 51&39*8= 5*:;*39= 574):.7*= )*8= 5.,2*398= S(7&3= 9*18= 6:*= )*8= (&749S34)*8= 4:= )*8= =&39-45->11*8= .1,*7= *9= &1 _`= +32.=b= /702&3= ]= 5*22.,8)&28`= +33/_= 8S&324.38`= (*79&.3*8= 2>(4-S9S749745-*8=(422*= *5-&1&39-*7&=&:89.3&* =*9= _=*=.,:& =8439=54:79&39='1&3(-*8=B.,_= -/`=(422*= 4349745& =855_=4:= :4>7.&= 855_=a=1*=(-&3,*2*39=)*=5.,2*39&9.43=3*89=)43(=5&8= *88*39.*1=A=1&=+.=&9.43=)*=1&=2>(4-S9S749745-.*_= = '&= 343885S(.+.(.9S= )*8= 5->1147(-.8 = &= 5&7= &.11*:78= 24397S= 6:*= 1&= 85S(.+.(.9S= )*8= (-&25.,3438= 2>(47-.?.*38= 3S9&.9= 5&8= 343= 51:8= 3S(S88&.7*= A= 1&= 2>(4-S9S749745-.*_= !3= (43(1:8.43`=1&=5488.'.1.9S=)*++*(9:*7=1&=97&38.9.43=)*=1&=2.=49745-.*=A=1&=2>(4-S9S749745-.*= )S5*3)7&= 51:8= )&)&59&9.438= (422*= 1&= 2475-414,.*= 4:= 1&= 5->8.414,.*= )*= 1&= +*:.11*_= 5&.11*:78`= 1&= 9&.11*= )*8= +*:.11*8= )*8= 2.=49745-*8$= *89= 84:;*39= 7S):.9*`= c*79&.3*8= >741*8= 2.=49745-*8$= 439= )*8= ;&7.&398= &5->11*8= (+_= >741&= 5.(9& `= Z.22*7= *9= &1 _`= ,**1= *9= )*= 342'7*:== 5.5&(9.8 =439=)*8=+*:.11*8=7S):.9*8`=(422*= _=2.(745->11& =*1488*= *9=&1 _`=,**.'_= 'S;41:9.43= )*= 1&= 2>(4-S9S749745-.*= 54:77&.9= &1748= 7S8:19*7= ):3*= (4S;:1:9.43= *397*= 1*= )*,7S=)-S9S749745-.*=*9=1&=9&.11*=)*8=+*:.11*8=*9=1*=342'7*=)*=8942&9*8=a=51:8=:3*=47(-.)S*= *89=-S9S749745-*`=24.38=8*8=+*:.11*8=8439=:9.1*8=54:7=8&=3:97.9.43=*9=51:8=1*8=57*88.438=)*= 8S1*(9.43=6:.=2&.39.*33*39=8*8=+*:.11*8=*9=1&55&7*.1=5-4948>39-S9.6:*=8*7&.*39=7*1B(-S*8`=(*= 6:.=5*:9=(43):.7*=A=1&=7S):(9.43=)*8=+*:.11*8=*9=1&=5*79*=)*=1&=5-4948>39-T8*_=

84 &= '=

(= 88.2.1&9.43=5-4948>39-S9.6:*== 241=1% ,_2 8, _8 8+ = 0

8, .24)47:2= _=)&2&843.:2= _=)&2&843.:2= 8- &'479.;:2= &1'.34= ;*79======Figure 36. .24)47:2=&'479.;:2 =&`=5-494=_8 _=B&'7*=*9=843=&.7*=)*=7S5&79.9.43=*3=,7.8S='_=1*99*= *85T(*=*89=&884(.S*=85S(.+.6:*2*39=A=)*8=$:88:1&(S*8=(1&)*=)*= -:88:1&=)*1.(&= 8:794:9`=*9='.*3=6:*= 5-4948>39-S9.6:*`= *11*= 7*85.7*= A= 1&= 1:2.T7*_= (= 1425&7&.843= )*= 1&= 7S5438*= A= 1&= 1:2.T7*= )*= _= &'479.;:2 `= )&57T8= .71&3)&= *9= &1_` = ,**0= *9= _= )&2&843.:2 = &1'.34= *9= ;*79_= !3= ,7.8=a= 4'8(:7.9S`= *9= *3= '1&3(=*=548.9.43=A=$=-=.**=2241_2 8, _8 8+= )433S*8=)*=1 &79.(1*= _=

85 .24)47:2= &'479.;:2 = *89= :3*= *85T(*= 2.=49745-*$= 2S).9*77&3S*33*= 5&79.(:1.T7*2*39=-S9S749745-*=*9=6:.=5488T)*=)*8=+*:.11*8=7S):.9*8=*9=97T8=(4147S*8=B.,_=-0_= !11*=*89=(&5&'1*=)*++*(9:*7=1&=5-4948>39-T8*=2&.8=*3=2.1.*:=3&9:7*1`=(*99*=*85T(*=8*=974:;*= 94:/4:78= *3= )*884:8= ):= 54.39= )*= (425*38&9.43_= !11*= 7*85.7*= )43(= 51:8= 6:*11*= 3*= 5-4948>39-S9.8*= *9= (*= 2&36:*= )*= (&7'43*= *89= &= 57.47. = (425*38S= 5&7= 1&55479= )*8= (-&25.,3438= B.,_= -0=b= .71&3)&= *9= &1_` = ,**0_= 1422*= 8&= 5-4948>39-T8*= 3*89= 5&8= &88*?= *++.(&(*= 54:7= &88:7*7= 8&= 57457*= 3:97.9.43`= *11*= *89= 5&7+4.8= (438.)S7S*= (422*= :3*= *85T(*= 2>(4-S9S749745-*= '*&0*`= +33.`= 4:= (422*= :3*= 2.=49745-*$= +&.'1*2*39= 5-4948>39-S9.6:*_=4:79&39`=(*99*=*85T(*=54:88*=)&38=)*8=2.1.*:==7*1&9.;*2*39=4:;*798`=*3= ,&77.,:*= 5&7= *=*251*= 4:73.S7.&8= *9= &1_` = ,**/_= !3= (438*7;&39= :3= &55&7*.1= 5-4948>39-S9.6:*=+43(9.433*1=*9=)*=5*9.9*8=+*:.11*8`=(*99*=*85T(*=&99*.39=:3=)*,7S=*=97U2*= )*=2.=49745-.*`=574(-*=)*=1&=2>(4-S9S749745-.*`=94:9=*3=(4143.8&39=:3=2.1.*:=8*(= &=57.47. = )S+&;47&'1*= &:== 2>(4-S9S749745-*8= 6:*= (4389.9:*= 1*= 2.1.*:= 2S).9*77&3S*3_= 5*= 2U2*= 47&1147-.?&= 97.+.)& = *89= :3*= 47(-.)S*= )*= 84:88'4.8= 8&38= +*:.11*8`= (438.)S7S*= (422*= 2>(4-S9S749745-*='*&0*`=+33.`=2&1,7S=1&=(4:1*:7=;*79*=)*=8&=9.,*_=!11*=5488T)*=:3*=(45.*= 3472&1*=):=,T3*=)*=1&=$:'.8(4=$.':148*=+`08'.85-485-&9*=4=>)&8*`=*3?>2*=7*85438&'1*= )*= 1&= 8>39-T8*= )&2.)43= *9= (439.*39= )*= 1&= (-14745->11*= &7*99= ]= B7*:)*389*.3`= ,**2=b= Z.22*7= *9= &1_` = ,**2_= &= 8.,3&9:7*= .84945.6:*= *89= 574(-*= )*= (*11*= )*8= 2>(4-S9S749745-*8= 3438;*79*8=2&.8=51:8=5&:;7*`=(*=6:.=8:,,T7*=6:*11*=*++*(9:*=1&=5-4948>39-T8*=Z.22*7= *9= &1_` =,**2_=1422*= _=&'479.;:2 `=(*99*=*85T(*=7*57S8*39*=:3=(&8=*=97U2*=)*=2.=49745-.*`=41= 1-S9S749745-.*= *89= 574(-*= )*= (*11*= )*8= 2>(4-S9S749745-*8`= 2&.8= 41= 1&= +43(9.43= 5-4948>39-S9.6:*=*89=(438*7;S*_== =1.389&7=)*=(*79&.3*8=51&39*8=5&7&8.9*8=6:.=:9.1.8*39=1&=5-4948>39-T8*=54:7=)&:97*8= +43(9.438= 6:*= 1&88.2.1&9.43`= (422*= 1&= '.48>39-T8*= )*8= 1.5.)*8= )&38= 1*8= ,7&.3*8= (-<*3)*7= *9= &1_` = ,**.=b= A7&:8*`= ,**2`= 4:= 1*= 894(0&,*= )&2.)43`= .1= *89= 574'&'1*= 6:*= 1&= 5-4948>39-T8*=84.9=2&.39*3:*=(-*?= _=&'479.;:2 =54:7=)&:97*8=+43(9.438=6:*=1&88.2.1&9.43_= '*8= &1'.348= 8*= 8439= 7S;S1S8= 8*38*.'1*8= &:== 5-S342T3*8= )*= 8:7(-&:++*= *9= 5&72.= 1*8= 2S(&3.82*8= 6:.= 5*72*99*39= )*= 7S,:1*7= 1&= 9*25S7&9:7*= 84:8= 1*++*9= ):3= 897*88= 1:2.3*:=`= (*79&.38= +439= .39*7;*3.7= 1&= $:'.8(4= ).88.5&9.43= )*= 1S3*7,.*= 5&7= 1&= 5-4947*85.7&9.43`= 29-47`=+33/=4:=1*8=51&89*8=7S,:1&9.43=)*=14:;*79:7*=)*8=8942&9*8=5&7=1&=1:2.T7*=74:,*`= 29-47`=+33/_=1*8=*=*251*8=24397*39=6:*=1S;41:9.43=)*=1&=2>(4-S9S749745-.*=3*=8*89= 574'&'1*2*39= 5&8= &((425&,3S*= ):3*= 5*79*= '7:9&1*= )*= 1&= 5-4948>39-T8*_= '&= 97&38.9.43= 2.=49745-.*82>(4-S9S749745-.*=&55&7&9=(422*=:3=(439.3::2=B.,_=-1=41=1&=57*88.43=)*= 8S1*(9.43=8:7=1&=5-4948>39-T8*=8*7&.9=1*39*2*39=7*1B(-S*_= . =49745-*8 = >(4-S9S749745-*8 = &:949745-.* = -S9S749745-.* =

9&.11*=)*8=+*:.11*8 `= 342'7*=)*=8942&9*8 = )472&3(* =

5S+*38*8=(4397*=1*8=-*7'.;47*8== = 749*(9.43=(4397*=1*8=897*88= 1:2.3*:==.39*38*8= =

5S(.+.(.9S=)&884(.&9.43=;&7.&'1* = #.1.*:=4:;*79 = &'.9&9=+47*89.*7 = = B.,:7*= -1_= 1439.3::2= *397*= 2.=49745-.*= *9= 2>(4-S9S749745-.*= *9= &)&59&9.438= 5488.'1*8= )S;*1455S*8=5&7=1*8=2>(4-S9S749745-*8_=

86 = 7&.3*8 =8&38=7S8*7;* =

*72.3&9.43= *72.3&9.43=2>(4-S9S749745-*8= 2>(4-S9S749745-*8=&;*(== &;*(=)*8=34387-.?4(943.&=a= )*8=7-.?4(943.&= (-&25.,3438=*(942>(47-.?.*38`= 8&5745->9*8=4:=5&7&8.9*8= =

-&8*=&):19*= -&8*=&):19*=&;*(=)*8=34387-.?4(943.&=a= &;*(=7-.?4(943.&= (-&25.,3438=*(942>(47-.?.*38`= 8&5745->9*8=4:=5&7&8.9*8

&345S* = 4:8 8'4.8 =

5.5->9*8 = :949745-*8 =

.1. *: = 4:;*79 =

41 =

:949745-*= = #6 = = #6 = # = = Figure 38. 5.;*78.9S= )*8= 7*1&9.438= 47(-.)S*88(-&25.,3438= )*= 1&= ,7&.3*= A= 1S9&9= &):19*= *9= 8*143= 1-&'.9&9_=

87 /_- 5*=1&:949745-.*=A=1&=2.=49745-.*= = '&= +&2.11*= )*8= 47(-.)S*8= *89=*=97U2*2*39= ).;*78.+.S*=&;*(= 57T8= )*= -*=***= *85T(*8= 17.''= *9=&1_` =,**-`=)439=33=8439=(-14745->1.*33*8`=*9=1*:78=2>(47-.?*8=8439=*11*8=&:88.= )*8=51:8=).;*78*8_=8S&324.38`=43=5*:9=84:1.,3*7=:3=97&.9=(422:3=a=94:9*8=1*8=47(-.)S*8= 8439= )S5*3)&39*8= )*= (-&25.,3438= 51:8= 4:= 24.38= 85S(.+.6:*8= &:= 242*39= )*= 1&= ,*72.3&9.43`=(*=(-&25.,343=&55479*=)*=1*&:=*9=):=(&7'43*=47,&3.6:*=A=1&=,7&.3*_=.4:9*8= 1*8= 47(-.)S*8= 8439= )43(= 2>(4-S9S749745-*8= &:= 24.38= 1478= )*= 1*:7= ,*72.3&9.43_= = 1B,*= &):19*`=1&=51:5&79=)*8=47(-.)S*8=;*79*8=8439=&884(.S*8=51:8=4:=24.38=85S(.+.6:*2*39=A=)*8= 7-.?4(943.&`= A= 6:.= *11*8= +4:73.88*39= )*8= 8:(7*8= .88:8= )*= 1*:7= 5-4948>39-T8*= *9= )439= *11*8= 7*O4.;*39=)*=1*&:=*9=)*8=8*18=2.3S7&:=_=5&57T8=1*8=7S(*39*8=S9:)*8=)*= .44)>*7&=7*5*38 `='*8= 7-.?4(943.&$=5*:;*39=3S&324.38=+4:73.7=A=147(-.)S*=):=(&7'43*=47,&3.6:*`=*9=1*8=+1:==)*= (&7'43*=+43(9.433*39=)43(=)&38=1*8=)*:==8*38=B.,_=-3`=1&2*743= *9=&1 _`=,**0=b=1&2*743= *9= &1 _`= ,**2_= 8S&324.38= 1*= +1:== 3*9= )*= (&7'43*= 47,&3.6:*= 8*++*(9:*= )*= 147(-.)S*= ;*78= 1*8= 7-.?4(943.&$=<43-843= *9=&1 _`=,**,_= =

:&39.9S=)*= +. =)S9*(9S*

1&7'43*= +*** 47,&3.6:* !&:`=8*18= 2.3S7&:=`=8`= +**

.,*=&(.3*8= &(.3*8= >(S1.:2

(-&25.,343 47(-.)S* 47(-.)S* (-&25.,343

7S8:19&9

'.1&3 47(-.)S* (-&25.,343 = = Figure 39. M.8*= *3= S;.)*3(*= ):= )4:'1*= 8*38= )*8= S(-&3,*8= )*= 1&7'43*= *397*= 147(-.)S*= *9= 1*= (-&25.,343=(-*?= .44)>*7&=7*5*38 `=:3*=47(-.)S*=9*77*897*=(-14745->111.*33*=)&57T8=1&2*743= *9=&1_` = ,**0_= = !3=7*;&3(-*`=1*8=47(-.)S*8=2.=49745-*8$=*9=2>(4-S9S749745-*8=3*=+4:73.88*39=5&8= )*= (&7'43*= 47,&3.6:*= A= 1*:78= (-&25.,3438= A= 1B,*= &):19*`= 8439= &884(.S8= A= )*8= (-&25.,3438= ).++S7*398= )*8= 7-.?4(943.&$a= )*8= (-&25.,3438= *(942>(47-.?.*38`= 8&5745->9*8=4:=5&7&8.9*8_=1*8=47(-.)S*8=;.;*39=94:9*8=*3=84:8='4.8=*9=(*99*=&884(.&9.43=A= )*8=(-&25.,3438=6:.=1*:7=+4:73.88*39=):=(&7'43*=47,&3.6:*=A=1B,*=&):19*=*89=(438.)S7S*= (422*=:3*=&)&59&9.43=A=1&=(4143.8&9.43=):=2.1.*:=+47*89.*7=6:.=5*72*9=)*=(425*38*7=1*= 2&36:*=)*=1:2.T7*=*3=84:88'4.8=*1488*= *9=&1_` =,**0_=:9&39=1&=1.2.9*=*397*=2.=49745-.*=*9= 2>(4-S9S749745-.*=8*2'1*=+14:*=*9=5*:9=U97*=.39*757U9S*=(422*=:3=(439.3::2`=&:9&39=1&= ).89.3(9.43=&;*(=1*8=&:97*8=47(-.)S*8=&:949745-*8=8*2'1*=51:8=2&76:S*_= =

88 ======

======B.,:7*=.*_= 75-7>8=8(4145&= =&1'.348=&:=3.;*&:=)*8=+*:.11*8_=-494=_=5:7'.3_=

89 4:79&39`=1*=+43(9.433*2*39=)*=1&884(.&9.43=3*=8*2'1*=5&8=8.=).++S7*39`=5:.86:*=1*8= +1:== )*= (&7'43*= 47,&3.6:*= ):= (-&25.,343= ;*78= 147(-.)S*= 8439= 5488.'1*8= &;*(= 1*8= 7-.?4(943.&`=1478=)*=1&=,*72.3&9.43=*9=A=1B,*=&):19*=B.,_=-3_=5&:97*=5&79`=)*8=&1'.348=439= S9S= 4'8*7;S8= *3= )*-478= )*8= 3*499.S*8`= )&38= )*8= 84:8897.':8= 6:.= 3*= (425479*39= 5&8= )*= 2>(4-S9S749745-*8= .&'1*&:= +._= 1*8= *85T(*8= 3*= 8439= 5&8= &884(.S*8= A= )*8= (-&25.,3438= *(942>(47-.?.*38= 7*+= 2&.8= A= )*8= 7-.?4(943.&`= 3S&324.38= )*8= .3).;.):8= &1'.348= 439= S9S= 4'8*7;S8`= )439= (*79&.38= *3= +1*:7= B.,_= .*`= (*= 6:.= 24397*= 6:.18= 8:7;.;*39= &:8)*1A= )*= 1&= ,*72.3&9.43_= 7.':=(439*3&39=)*8= 85T(*= &'.9&9= &>8= 2>(4-S9S749745-*8= >2'.).:2=85_= 1-.3*= %:.= 47&1147-.?&=97.+.)&= 4:88'4.8= B7&3(* = 1&9&39-*7&=2.347= <&5&3= 843= = = = = 75-7>8=8(4145&== 7_=4'&*8&== 7_=85*(:1:2= B7&3(* = 843= #.1.*:=4:;*79 = 7_-4148*7.(&= = = 7_.38*(9.+*7&= ,&(9>1477-.?&=8&2':(.3& = = &'1*&:=+._ =1&8=)&1'.3.82*=949&1=+*:.11*`=9.,*`=+1*:7=(-*?=)*8=*85T(*8=)47(-.)S*8=*3=)*-478=)*8= 3S499.S*8=(422:3.(&9.438=5*78433*11*8=)*=._=#&1.34;&`= _=(&55&9.((.=`=._=&:0&<&_=

1*8= &1'.348= 8:7;.;*39= 574'&'1*2*39= *3= S9&39= 2>(4-S9S749745-*8= 94:9= *3= S9&39= &884(.S=A=)*8=7-.?4(943.&_='*8=.3).;.):8=;*798=54:77&.*39=574'&'1*2*39=U97*=2.=49745-*8`= *3= S9&39= &884(.S8= A= )*8= 7-.?4(943.&_= 5&38= (*= (&8`= 1&= 1.2.9*= *397*= 1*8= 47(-.)S*8= ).9*8= &:949745-*8= &848(.S*8= A= )*8= 7-.?4(943.&$= *9= 1*8= 47-(.)S*8= 2.=49745-*8$= *9= 2>(4-S9S749745-*8= &55&7&9= &:88.= +14:*_= 5&.11*:78`= (*79&.38= (1&)*8= )*= 7-.?4(943.&$= *7&94'&8.).:2`= #*'&(.3& = 8*= 8439= 7S;S1S8= *(942>(47-.?.*38`= *9= &884(.S8= A= )*8= A= )*8= 47(-.)S*8= 2>(4-S9S749745-*8= *3(&)7S= -=b= F*.88= *9= &1 _`= ,***=b= &&,&2*= *9= &1 _`= ,**2'_= 1*8= (1&)*8= 8439= ).89.3(98= )*8= (1&)*8= )*= 7-.?4(943.&$= &884(.S8= &:== 47(-.)S*8= ;*79*8= 2&.8= 1&= 574=.2.9S=9&=4342.6:*=5*:9=8&38=)4:9*=+&(.1.9*7=1&55&7.9.43=)*=1&=2.=49745-.*=4:=)*=1&= 2>(4-S9S749745-.*_= '&55&7.9.43= )*= 1&= 2.=49745-.*= 3*89= )43(= 5&8= 2&76:S*= 5&7= :3= (-&3,*2*39= )*= +43(9.433*2*39=)*=1&884(.&9.43`=5:.86:*=94:9*8=1*8=47(-.)S*8=5*:;*39=7*(*;4.7=):=(&7'43*= 47,&3.6:*= )*= 1*:78= (-&25.,3438= 2>(47-.?.*38`= 2&.8= 5&7= :3= (-&3,*2*39= .25479&39= )*= 1S(414,.*= ):= 5&79*3&.7*= 2>(47-.?.*3_= 1*8= 34:;*&:== (-&25.,3438= 2>(47-.?.*38`= 6:.18= 84.*39=*(942>(47-.?.*38`=5&7&8.9*8=4:=8&5745->9*8=2&.8=5&8=7-.?4(943.&$=8&545->9*=8439= 8&38= )4:9*= (&5&'1*= )&55479*7= :3*= 51:8= ,7&3)*= 6:&39.9S= )*= (&7'43*= 47,&3.6:*= &:== 47(-.)S*8=&):19*8_=1*99*=->549-T8*=*89=;&1&'1*=54:7=1*8=*(942>(47-.?.*38=.&>147=]=7:38`= +331`=6:.=7*O4.;*39=+*=):=(&7'43*=574):.9=5&7=1*8=&7'7*8=.2&7)=]=5:7&11`=,**._=03= )*8= (-&25.,3438= 5&7&8.9*= &884(.S= &:== 2>(4-S9S749745-*8`= 72.11&7.&= 2*11*&= .*7&8-.9&`= +32/`=:3=:3=5&7&8.9*=7*)4:9&'1*=6:.=)S(.2*=)*8=+47U98=*39.T7*8=$*)+*73`=+312=*9=&((T)*= )43(=A=)*=,7&3)*8=6:&39.9S8=)*=(&7'43*=47,&3.6:*_=1*=51:8=,7&3)=&55479=)*=(&7'43*=5&7= 1*8= (-&25.,3438= 2>(47-.?.*38= 5*72*997&.9= &:== 47(-.)S*8= )*= )*;*3.7= &88*?= +&(.1*2*39= 2.=49745-*8`=;4.7*=2>(4-S9S749745-*8_=

90 5S(.&1.89*=

,*72.3&9.43=1.2.9S*=5&7=1&= 3S(S88.9S=)*=7*3(4397*7=1*= &'.9&9=*924:=+147&.843== B147&.843=)*=(4:79*=):7S* :3=(-&25.,343= 5*3)&39=:3*=8&.843=5*:== 2>(47-.?.*3=(425&9.'1* +&;47&'1*=&:=5411.3.8&9*:7

B&.'1*=)*38.9S= !++*9=11**=+479 S1*(9.43=)*== 5411.3.8&9*:78=,S3S7&1.89*8= 4:=&:94,&2.*

= = Figure 41. B&(9*:78=+&;47.8&39=1&:94,&2.*=4:=1&=5411.3.8&9.43=5&7=)*8=.38*(9*8=,S3S7&1.89*8=(-*?=1*8= 2>(-4-S9S749745-*8_='*8=+1T(-*8=.3).6:*39=1*8=(438S6:*3(*8=5488.'1*8_=

91 /_. '*=7S,.2*=)*=7*574):(9.43=)*8=2>(4-S9S749745-*8= = 84:8=&;438=S9:).S=1&=7*1&9.43=*397*=1*8=(-&25.,3438=*9=1*8=47(-.)S*8=2&.8=>&898.1= :3=1.*3=&;*(=1&=7*1&9.43=47(-.)S*85411.3.8&9*:7==1425&7S=A=1*:78=574(-*8=5&7*398=;*798`=1*8= 51&39*8= 2>(4-S9S749745-*8`= 5&8= 8*:1*2*39= 1*8= 47(-.)S*8= 2>(4-S9S749745-*8`= 574):.8*39= )*8=9.,*8=+1*:7.*8=6:.=&55&7&.88*39=+:,&(*2*39=*3=)*-478=):=841`=4:=5*:;*39=2U2*=+1*:7.7= 84:8= 9*77*= :22*7-&>*8`= +3/+_= '*8= 9.,*8= 8439= 51:99= 7S):.9*8`= 5479*39= 84:;*39= 5*:= )*= +1*:78`=*9=1*8=+1*:78=439=84:;*39=1&=2U2*=(4:1*:7=6:*=1&=9.,*_=1*8=*85T(*8=439=5&7=&.11*:78= 9*3)&3(*= A= U97*= &:94,&2*8= 4:= '.*3= A= U97*= 5411.3.8S*8= 5&7= )*8= 5411.3.8&9*:78= ,S3S7&1.89*8= '*&0*`=+33._=5&57T8=2*8=*=5S7.*3(*8=2*3S*8=8:7= _=2439&3& `=(*99*=*85T(*=*89=&:94,&2*= (+=5&7&,7&5-*=-_/_=!3=7*;&3(-*`= _=*=.,:& =8*2'1*=51:99=&114,&2*=2&.8=1*8=5411.3.8&9*:78= 3439=5&8=S9S=.)*39.+.S8= (+_ =&79.(1*= _= 1= 8*2'1*= &.38.= 6:*= 1*=97U2*= 85S(.&1.8&9.43= ;.88A8;.8= )*= (-&25.,3438= 8&((425&,3*=):3=7*1B(-*2*39=)*=1&=85S(.+.(.9S=;.88A8;.8=)*8=5411.3.8&9*:78`=;4.7*=):3*= .3)S5*3)&3(*= ;.88A8;.8= )*:== &:94,&2.*_= F&9*72&3= *9= .)&7943)4= ,**2= 84:1.,3*39= 6::3*= 85S(.&1.8&9.43= ;.88A8;.8= )*8= )*:== 8>2'.48*8= 1.2.9*7&.9= 9*11*2*39= 1S(414,.*= )*= (*8= 47(-.)S*8=6:.1=3*=54:77&.9=8&,.7=):3*=897&9S,.*=S;41:9.;*2*39=89&'1*_=1*5*3)&39`=.18=3*= ).8(:9*39=5&8=51:8=1*=7S,.2*=)*=7*574):(9.43=)*8=2>(4-S9S749745-*8`=84:;*39=&:94,&2*8_= .4:9= )&'47)`= 1&= 51:5&79= )*8= 2>(4-S9S749745-*8= 8439= 85S(.+.6:*8`= 1*8= (4397&.39*8= 54:7=974:;*7=1*:7=5&79*3&.7*=2>(47-.?.*3=8439=)43(=97T8=+479*8=*9=1.2.9*39=1*:7=,*72.3&9.43= )&38= )*8= 2.1.*:== 41= 1*= (-&25.,343= 3*89= 5&8= 57S8*39`= 8&:+= 8.= 1*8= (-&25.,3438= 2>(47->?.*38= (425&9.'1*8= 8439= 97T8= (422:38`= (422*= 1*8= 7:88:1&(S*8= *9= 1*8= 9-S1S5-47&(S*8_= #U2*= )&38= 1*8= 2.1.*:== +&;47&'1*8`= 1&= )*38.9S= )*8= 9.,*8= &S7.*33*8= ):3*= *85T(*= 2>(4-S9S749745-*= 7*89*= +&.'1*_= '&= 7*574):(9.43= 5&7= &114,&2.*= .251.6:&39= 1*8= 5411.3.8&9*:78=*89=)S5*3)&39*=)*=1&=)*38.9S=)*=1&=545:1&9.43=*9=*89=51:8=(429*:8*=)&38=1*8= 5*9.9*8=545:1&9.438=6:.=3*=5*:;*39=5&8=&88:7*7=1*:7=7*574):(9.43_=1*99*=9*3)&3(*`=:3=(&8= )*++*9=11**= 9*5-*38=]= :9-*71&3)8`= +333=b= 5473.*7= *9= &1_` = ,**2`=3*= 8&551.6:*= 5&8= &:== 5*9.9*8= 545:1&9.438= )*= 51&39*8= &:94,&2*8`= )439= 1&= 7*574):(9.43= *89= .3)S5*3)&39*= )*= 1&= )*38.9S_= 1*(.= &= 574'&'1*2*39= (43):.9= A= (4397*88S1*(9.433*7= 1&114,&2.*= (-*?= 1*8= *85T(*8= 2>(4-S9S749745-*8=4:=A=8S1*(9.433*7=)*8=5411.3.8&9*:78=,S3S7&1.89*8=&88*?=(422:38=54:7= &88:7*7=1&=7*574):(9.43=2U2*=)&38=)*8=545:1&9.438=)*=+&.'1*=)*38.9S=B.,_=.+_= '&=)*38.9S=3*89=(*5*3)&39=5&8=1&=8*:1*=*=51.(&9.43=5488.'1*_='*&0*=+33.=*=51.6:*= (*99*=5411.3.8&9.43=5&7=)*8=.38*(9*8=,S3S7&1.89*8=(&7=(*8=2>(4-S9S749745-*8=8439=5*9.9*8`=3*= )S5&88*39= 5&8= 1&= (4:(-*= -*7'&(S*= *9= 8439= &88*?= 5*:= ;.8.'1*8_= 5*= 51:8`= 1*8= +1*:78= 8*= 8.9:*7&.*39=)&38=)*8=84:88'4.8=41=8*=974:;*39=57.3(.5&1*2*39=)*8=.38*(9*8=,S3S7&1.89*8_=&7= &.11*:78`= .1= *89= +7S6:*39= )*= 974:;*7= 51:8.*:78= *85T(*8= 2>(4-S9S749745-*8= ).++S7*39*8= &:= 2U2*=*3)74.9`=6:*=(*=84.9=)*8=47(-.)S*8=*9=*7.(&(S*8=*3=7S,.43=9*25S7S*=4:=)*8=47(-.)S*8`= ':72&33.&(S*8`= ,*39.&3&(S*8= *9= 97.:7.)&(S*8= *3= 7S,.43= 9745.(&1*= '*&0*`= +33._= !3= .-&1&3)*`=(*8=*85T(*8=+1*:7.88*39=84:;*39=A=1&=8&.843=)*8=51:.*8`=&:=(4397&.7*=)*8=&:97*8= 47(-.)S*8=S5.5->9*8=6:.=+1*:7.88*39=51:99=*3=)S':9=)*=8&.843=8T(-*=*.)*3+&)*3`=+312_=1= *89=574'&'1*=6:*=1&=8&.843=*9=1*=2.1.*:`=(422:3=A=94:9*8=(*8=2>(4-S9S749745-*8`=3*=84.*39= 5&8=+&;47&'1*8=&:==5411.3.8&9*:78`=2&.8=51:99=+&;47&'1*8=&:=2S9&'41.82*=)*8=&:949745-*8= *9=)*8=(-&25.,3438=94:8=1*8=(&7545-47*8=8439=;.8.'1*8=*3=8:7+&(*=A=1&=8&.843=)*8=51:.*8`=*9= 1&= ;S,S9&9.43= *89= 1:=:7.&39*_= 1*= )S(&1&,*= )*= 1&= 5-S3414,.*= )*8= 47(-.)S*8= 2>(4-S9S749745-*8=9745.(&1*8=6:.=8:.;7&.9=51:99=(*11*=)*8=&7'7*8=*9=)*8=(-&25.,3438=6:*= (*11*=)*8=5411.3.8&9*:78`=54:77&.9=+&;47.8*7=&:88.=1&:94,&2.*=B.,_=.+_=

92 ======&=

+=(2 =

'= (=

+=22 = +=22 ======Figure 42. #475-414,.*= ):= 8>89T2*= 7&(.3&.7*= )*= 5.54,.:2= &5->11:2 = &= *9= 2.8*= *3= S;.)*3(*= )*8= 8941438=*9=)*=1*:78=':1'.11*8=+1T(-*8='`=6:.=)S;*1455*39=)*8=7-.?4)*8=5&7=1&=8:.9*=(`=5-4948=1_= *.3?`= (+_= &79.(1*= _=

93 1*= 5-S342T3*= 3*89= 5&8= 4'8*7;S= *3= 2.1.*:= 9*25S7S`= 41= 1*8= 2>(4-S9S749745-*8= +1*:7.88*39=*3=S9S_=!9433&22*39`= _=*=.,:& `=6:.=8*=7&5574(-*=51:8=)*8=24)T1*8=9*25S7S8`= *89= &114,&2*= *9= +1*:7.9= A= 1&= 8&.843= 8T(-*`= *3= &;7.1`= 2&.8= )&38= )*8= 2.1.*:== )&19.9:)*= 7*1&9.;*2*39=-:2.)*8=94:9*=1&33S*= (+_ &79.(1*= _=5&38=1&=2*8:7*=41=1&=8S(-*7*88*=8*2'1*= )S+&;47&'1*=&:==2>(4-S9S749745-*8=)&57T8=1S9:)*=)*8=.3).;.):8=&1'.348= (+_ &79.(1*= `=.1= *89= &:88.= 5488.'1*= 6:*= 1*= )S(&1&,*= )*= 5-S3414,.*= )*8= 2>(4-S9S749745-*8= 9745.(&1*8= 84.9= (4397&.39=5&7=(*99*=8&.843=8T(-*`=6:.=*89=5&7=&.11*:78=51:8=+&;47&'1*=&:==5411.3.8&9*:78=)*8= &:97*8=47(-.)S*8_= !3+.3= 1&= ):7S*= )*= ;.*= )*8= +1*:78= )*8= 2>(4-S9S749745-*8= *89= (4:79*=a= _= 2439&3& = 574):.9= )*8= +1*:78`= )*8= +7:.98= *9= ).85*78*= 8*8= ,7&.3*8= *3= A= 5*.3*= :3= 24.8= #_= $4>= ]= _= F&99-&3&`= 4'8*7;&9.43= 5*78433*11*= 9&3).8= 6:.1= +&:9= (4259*7= *3;.743= ,= 24.8= (-*?= _= *=.,:& = (+_ =&79.(1*= =*9=6:*=1*8=47(-.)S*8=S5.5->9*8=8439=57.8S*8=)*8=+1*:7.89*8=54:7=1*:7= 143,:*=):7S*=)*=+147&.843_=1*99*=7S):(9.43=)*=1&=+147&.843=;&=)*=5&.7=&;*(=1&=7S):(9.43=)*= 1&55&7*.1= ;S,S9&9.+= &S7.*3= '*&0*`= +33.= *9= 1&= 57S)42.3&3(*= )*= 1&= ;.*= 84:9*77&.3*_= '*8= 2>(4-S9S749745-*8=.3;*89.88*39=)43(=)*=+&O43=242*39&3S*=)&38=1&=7*574):(9.43=8*=:S*`= *9=(*99*=7*574):(9.43=*89=)&:9&39=51:8=*++.(&(*=6:*11*=*89=&:94,&2*=4:=6:*11*=8*=+&.9=5&7=1*= '.&.8=)*=5411.3.8&9*:78=,S3S7&1.89*8_= %3= 5*:9= )43(= .39*757S9*7= (*99*= 9*3)&3(*= A= 1&:94,&2.*= 4:= &:= 7*(7:9*2*39= )*= 5411.3.8&9*:78=,S3S7&1.89*8=(422*=:3*=7S8:19&39*=)*=(4397&.39*8=8:7=1-&'.9&9`=1&=)*38.9S`=*9= 1.3;*89.88*2*39= )&38= :3*= 7*574):(9.43=7S):.9*= )&38= 1*= 9*258=*9= )&38= 1*85&(*=B.,_= .+_= := 5&88&,*`= 1*= 7&5574(-*2*39= &;*(= )&:97*8= 5&7&8.9*8= 5*72*9= &:88.= )*= 84:1.,3*7= 1&:97*= 24)*=)*=7*574):(9.43=)*8=47(-.)S*8=2>(4-S9S749745-*8=a=1&=7*574):(9.43=&8*=:S*_=1*11*= ) 5.54,.:2= &5->11:2 = *89= 5&7= *=*251*= &88:7S*= 5&7= )*= 143,8= 8941438= 6:.= 5479*39= )*8= ':1'.11*8= B.,_= .,=b= (+_ = &79.(1*= =b= )*8= (&8= 8.2.1&.7*8= 8439= 4'8*7;S8= (-*?= )&:97*8= 51&39*8= 2>(4-S9S749745-*8=A1.2*84;&`=,**1_= !3=(43(1:8.43=43=5*:9=(4251S9*7=1*=24)T1*=)S;41:9.43=)*=1&=2>(4-S9S749745-.*=A= 5&79.7= )*= 1&= 2.=49745-.*= 5&7= (*8= S1S2*398= 8:7= 1*= 24)*= )*= 7*574):(9.43= B.,_= .-_ :949745-*8 .=49745-*8 >(4-S9S749745-*8

&:949745-.* -S9S749745-.*

9&.11*=)*8=+*:.11*8`== 342'7*=)*=8942&9*8 )472&3(*

5S+*38*8=(4397*=1*8=-*7'.;47*8==

749*(9.43=(4397*=1*8=897*88=1:2.3*:==.39*38*8=

5S(.+.(.9S=)&884(.&9.43=;&7.&'1*

#.1.*:=4:;*79 &'.9&9=+47*89.*7 :94,&2.*=4:=5411.3.8&9*:78=,S3S7&1.89*8 = $*574):(9.43=&8*=:S*=84:9*77&.3* = 884(.&9.43=A=)*8=(-&25.,3438= 884(.&9.43=A=)*8=(-&25.,3438=(&5&'1*8=)*=+4:73.7= (&5&'1*8=)*=+4:73.7=8:++.8&22*39= 8:++.8&22*39=)*=(&7'43*=54:7=&88:7*7=94:9=4:=5&79.*=)*= )*=(&7'43*=54:7=1&=,*72.3&9.43=a== 1&=3:97.943=)*=1&=51&39*=&):19*=a== 7-.?4(943.& = *(942>(47-.?.*38`=5&7&8.9*8=4:=8&5745->9*8= = B.,:7*= .-_= '*= (439.3::2= &:949745-.*82>(4-S9S749745-.*= (-*?= 1*8= 47(-.)S*8= *9= &)&59&9.438= 5488.'1*8=A=1&=2>(4-S9S749745-.*=A=1B,*=&):19*_=

94 = = = = = B1:==)*= = -&25.,3438 =

8(42>(T9*8= 5.5&(9.8=-*11*'47.3* -S1S5-47&(S*8= 30 7-.?4(943.& = +** 5.5&(9.8=5&1:897.8

:88:1&(S*8= /- 5.5&(9.8=+1&;& -S1S5-47&(S*8= S'&(.3&1*8= 5->1147(-.8 1, 1&;:1.3&(S*8`=c=

.24)47:2 :88:1&(S*8 = 2. * *499.& S'&(.3&1*8 = 32 * .89*7&=2&(7&39-&

+** .89*7&=4;&9&

+** .89*7&=/&543.(&

-S1S5-47&(S*8= *5-&1&39-*7&=*7*(9& +** 479.3&7.&(S*8`=c=

0, *5-&1&39-*7&=8:'&5->11& -S1S5-47&(S*8 =

+** * *5-&1&39-*7&=+&1(&9& -S1S5-47&(S*8 =

*5-&1&39-*7&=143,.+41.& -S1S5-47&(S*8= 479.3&7.&(S*8`=c= "&1247(-.8

(-&25.,343=*(942>(47-.?.*3=;&8.).42>(T9*8= (-&25.,343=*(942>(47-.?.*3=8(42>(T9*8= 7-.?4(943.&

= Figure 44. ->14,S3.*= )*8=3*499.S*8= S9&'1.*= )&57T8= 1*8= .= *9= 2&9A= 5&7= 2&=.2:2= )*= &7(.243.*= &;*(=+***=7S5S9.9.438=(411&'47&9.43=&;*(=._=&:0&<&`=<&7).3='49&3.6:*=)*=.8:0:'&`=<&543`=+1:==)*= (&7'43*= *397*= (-&25.,3438= *9= 47(-.)S*8= *9= .)*39.9S= )*8= 5&79*3&.7*8= 2>(47-.?.*38_= '*8= *85T(*8= 2>(4-S9S749745-*8=8439=7*57S8*39S*8=*3=748*`=1*8=;*79*8=*3=34.7`=1*8='7&3(-*8=47&3,*8=(425479*39= )*8=2.=49745-*8$=*9=748*8=)*8=2>(4-S9S749745-*8_='&89S7.86:*=.3).6:*=1*8=(&8=)*=85S(.+.(.9S_=

95 /_/ '*8=3*499.S*8=*9=1&=2>(4-S9S749745-.*= = .=1&=2.=49745-.*=*9=1&884(.&9.43=A=)*8=7-.?4(943.&8$=*89=:3=(&7&(9T7*=&3(*897&1`='.*3= 5*:=)*=3*499.S*8=8439=*3=7*;&3(-*=&884(.S*8=A=)*8=7-.?4(943.&8_='*8='.89T7*8`=)*8=47(-.)S*8= +47*89.T7*8`= 8439= (433:*8= 54:7= 1*:78= ;&7.&9.438= 8&.8433.T7*8= )*= 1*:7= 2>(47-.?&9.43= B:72&3=]=.7&55*`=+31+=*9=7*89*39=:3=2>89T7*=(&7=1*:78=(-&25.,3438=2>(47-.?.*38=3439= 5&8= S9S= .)*39.+.S8_= 8S&324.38`= (*79&.3*8= 2*8:7*8= .84945.6:*8= *3= +439= )*8= 2.=49745-*8$= *'&:*7= ]= #*>*7`= ,**-_= !11*8= 54:77&.*39= 7*57S8*39*7= (*9= S9&9= &3(*897&1`= 8.= *11*8= 8439= &884(.S*8= A= )*8= 7-.?4(943.&8_= &7= &.11*:78= 43= 7*974:;*= 6:*16:*8= 7-.?4(943.&8$= )&38= 1*8= 7&(.3*8=)*= _=)&2&843.:2 =<:14:= *9=&1_` =,**/`= _=143,.+41.&= '&).*= *9=&1_` =,**0`=*9=2U2*= _= 2439&3& =(+_ &79.(1*= _= '*8= 3*499.S*8= 8439= 54:7= 1&= 51:5&79= 9*25S7S*8`= +47*89.T7*8= *9= &884(.S*8= A= )*8= (-&25.,3438= *(942>(47-.?.*38`= )439= *11*8= 7*O4.;*39= ):= (&7'43*_= '&884(.&9.43= &;*(= (*8= (-&25.,3438=*(942>(47-.?.*38=&=S9S=&(6:.8*=97T8=99=*9=5*72*9=:3*=2.=49745-.*='*&:(4:5= 51:8=,7&3)*`=;4.7*=:3*=2>(4-S9S749745-.*=(4251T9*=B.,_=.._='*=(&8=) 5.5&(9.8=5&1:897.8` = &884(.S*=A=)*8=7-.?4(943.&8$=*9=(4143.8&39=1*8=2.1.*:==-:2.)*8=5*:9=U97*=.39*757S9S=(422*= :3*= 5*79*= 8*(43)&.7*= )*= 1&884(.&9.43= &:== (-&25.,3438=*(942>(47-.?.*38=.)&7943)4= *9= &1 _`=,**.=b='&).*= *9=&1 _`=,**0_=.4:9*8=1*8=&:97*8= 5.5&(9.8 =8439=&884(.S*8=A=)*8=(-&25.,3438= *(942>(47-.?.*38= &8(42>(T9*8= *1488*= *9= &1_` = ,**.'=b= .)&7943)4= ]= $*&)`= ,**2_= 1*= (-&3,*2*39=)*=5&79*3&.7*8=2>(47-.?.*38=8(42>(T9*8=;8_ =&8.).42>(T9*8=5*:9=U97*=;:= (422*= :3*= (438S6:*3(*= ):3*= 343885S(.+.(.9S= &3(*897&1*= )&884(.&9.43= ;.88A8;.8= )*8= *(942>(47-.?.*38`=6:.=&:947.8*=)*=9*18=(-&3,*2*398_=1*99*=34:;*11*=&884(.&9.43=5*72*9=&:== 5.5&(9.8 =)*=(436:S7.7=)*=34:;*&:==2.1.*:=`=349&22*39=)*8=2.1.*:==51:8=4:;*798=4:=51:8= 5*79:7'S8`= 4:= 8*= 974:;*39= (*8= (-&25.,3438= *(942>(47-.?.*38= *9= )S;.9*7= 1&= (425S9.9.43= 54:7= 1*=514.9&9.43= )*8= 2U2*8= (-&25.,3438= &;*(= 1*8= *5-&1&39-*7& _= 1*99*= *=(1:8.43= (425S9.9.;*=54:77&.9=5*72*997*=A=)*8=)*:==,*37*8=)*=(4-&'.9*7=8&38=8:7*=514.9*7=1*=1#8_= 5*=51:8`=1*8= 5.5&(9.8 =439=(436:.8=)*=34:;*&:==-&'.9&98`=51:8=7:)S7&:==*9=51:8=3.9745-.1*8`= 41=1*8=(-&25.,3438=*(942>(47-.?.*38=&8(42>(T9*8=8*2'1*39=51:8=&'43)&398=B_=$.(-&7)= (422:3.(&9.43= 5*78433*11*_= &7= *=*251*`= 1*8= 5.5&(9.8 = 5*:;*39= (4143.8*7= )*8= ):3*8= %,:7&8.8:/.9&=]=&0&<&`=,**2&=4:=)*8=2&7&.8_= '*8= .3).;.):8= &1'.348= ) 5.5&(9.8 = *9= )*= *5-&1&39-*7& = 5*:;*39= 8:7;.;7*= )&38= 1&= 2*8:7*=41=1&55479=)*=(&7'43*=*89=(438S6:*39=(*=6:.=3*89=5&8=1*=(&8=&;*(=1*8=7-.?4(943.&8= *9= 8.18= 3*= +472*39= 5&8= )*85T(*= 2>(4-S9S749745-*`= .18= 24397*39= 3S&324.38= 6:*= 1&= 97&38.9.43=;*78=1&=2>(4-S9S749745-.*=*89=+&(.1.9S*=5&7=1&=2.=49745-.*=&;*(=)*8=(-&25.,3438= *(942>(47-.?.*38_= .4:9*8= 1*8= 3*499.S*8= 2>(4-S9S749745-*8= 8439= &884(.S*8= A= )*8= (-&25.,3438=*(942>(47-.?.*38`=*9=8*=8439=).;*78.+.S*8=)&38=)*8=2.1.*:==)42.3S8=5&7=)*8= &7'7*8= *(942>(47-.?S8`= 8&:+= :3= ,*37*`= &1247(-.8 `= 57S8*39= *3= 2S7.6:*= ):= :)= *9= 6:.1= 8*7&.9= .39S7*88&39= )S9:).*7`= )&:9&39= 51:8= 6:.1= 8*= 51&(*= A= 1&= 7&(.3*= )*= 1&7'7*= 5->14,S3S9.6:*=)*8=3*499.S*8_= '&= ).;*78.9S= )*8= &884(.&9.438`= 1&55&7.9.43= 7S(:77*39*= )*= 1S9&9= 2>(4-S9S749745-*= &1'.348= *9= 2>(4-S9S749745-*8`= 1*= 342'7*= .25479&39= )*= 2.=49745-*8$= *9= 1*8= 7S;*78.438= 5488.'1*8= 5.5&(9.8=5&1:897.8 =+439=)*=(*99*=97.':=:3=24)T1*=54:7=1S9:)*=)*=1S;41:9.43=)*=1&= 2>(4-S9S749745-.*_= 1=+&:)7&.9= 2&.39*3&39= 8.39S7*88*7= &:== 57*88.438= )*=8S1*(9.43= 8:7= 1*8= ,T3*8= )*= 1&= 5-4948>39-T8*= )&38= (*8= *85T(*8`= *9= 349&22*39= (-*?= 1*8= 2.=49745-*8`= &+.3= )S;&1:*7= 1*8= (438S6:*3(*8= ,S3S9.6:*8= )*= 1&= 2.=49745-.*`= )*= 1&= 2>(4-S9S749745-.*= *9= 54:;4.7=)433*7=:3*=).2*38.43=9*2547*11*=A=(*8=(-&3,*2*398_=

96 = = 14397*88S1*(9.43= &= = 55&7.9.43=*9=2&.39.*3=&:=3.;*&:=)*=1*85T(* ;&7.&'1*=(=)S7.;* = = 7*574):(9.43 + 2:9&9.43 (7>594541>2475-.82* = 2:9&9.43 = '= :949745-*8 .=49745-*8 >(4-S9S749745-*8 &:949745-.* -S9S749745-.* 9&.11*=)*8=+*:.11*8`== 342'7*=)*=8942&9*8 )472&3(*

5S+*38*8=(4397*=1*8=-*7'.;47*8== 749*(9.43=(4397*=1*8=897*88=1:2.3*:==.39*38*8=

5S(.+.(.9S=)&884(.&9.43=;&7.&'1*

#.1.*:=4:;*79 &'.9&9 =+47*89.*7 :94,&2.*=4:=5411.3.8&9*:78= $*574):(9.43=&8*=:S*= 884(.&9.43=A=)*8= 884(.&9.43=A=)*8=(-&25.,3438=(&5&'1*8=)*=+4:73.7= (-&25.,3438=(&5&'1*8=)*= 8:++.8&22*39=)*=(&7'43*=54:7=&88:7*7=94:9=4:= +4:73.7=8:++.8&22*39=)*= 5&79.*=)*=1&=3:97.943=)*=1&=51&39*=&):19*=a== (&7'43*=54:7=1&=,*72.3&9.43=a== *(942>(47-.?.*38`=5&7&8.9*8=4:=8&5745->9*8 7-.?4(943.& = = (=

= %7(-.) é*8 884(.&9.43= à= #.1.*: #.1.*:= )*8=(-&25.,3438 .*25 é7é 9745.(&1

= #>(4- é9é749745-*8 !(942>(47-.?.*38 ((( (((

&5745->9*8 8 (((

&7&8.9*8 8 (((

5 é(.+.6:*8 ((( (=4:= 8

#.=49745-* 8 5 é(.+.6:*8 (=4:=8 8

!(942>(47-.?.*38 ((( (

Figure 45. 'S;41:9.43= )*= 1&= 2>(4-S9S749745-.*=a= &`= 2S(&3.82*8= *=51.6:&39= 1*= 2&.39.*3= )*= 2>(4-S9S749745-*8= 543(9:*18= (-*?= )*8= *85T(*8= 2.=49745-*8$= b= '`= (439.3::2= &:949745-.*8 2>(4-S9S749745-.*= *9= 97&.98= 2475-414,.6:*8= )S;*1145S8= 5&7= 1*8= 2>(4-S9S749745-*8`= 1S9&9= 2.=49745-*8$= *89= 7S;*78.'1*= )&57T8= 1*= (&8= )!5.5&(9.8= 5&1:897.8=b= (`= ).;*78.9S= )*8= &884(.&9.438= 2>(47-.?.*33*8=)*8=2.=49745-*8$=*9=)*8=2>(4-S9S749745-*8=*3=2.1.*:=9*25S7S=*9=9745.(&1_=

97 Conclusions et perspectives = '&= 2>(4-S9S749745-.*= )*8= 47(-.)S*8= A= 1&= ,*72.3&9.43= &55&7&9= (422*= :3= 97&.9= &3(*897&1`=*9= (*99*=(&5&(.9S= A=7*(*;4.7= ):= (&7'43*= )*8= (-&25.,3438= 2>(47-.?.*38= +&(.1.9*= 574'&'1*2*39= 1&55&7.9.43= )*85T(*8= 2.=49745-*8`= ;4.7*= 2>(4-S9S749745-*8_= '*= 8(-S2&= )S;41:9.43= 574548S= B.,_= ./`= )*= 1&:949745-.*= A= 1&= 2>(4-S9S749745-.*= *89= :3= (-*2.3= S;41:9.+= 5488.'1*= 6:.= 2T3*= A= 1&55&7.9.43= )*85T(*8= 2>(4-S9S749745-*8_= 1*= (-*2.3= 5*:9= U97*=7*57S8*39S=5&7=:3=(439.3::2`=41=1*=54:7(*39&,*=)-S9S749745-.*=;&7.*=(439.3:*2*39_= 1*5*3)&39`= (*= (439.3::2= 3*89= 5&8= 47.*39S= *9= )*8= 7S;*78.438= 8439= 5488.'1*8`= (422*= 1*= 8:,,T7*=1*=(&8=):=(1&)*=) 5.5&(9.8=5&1:897.8 `=*9=8.=1*=54:7(*39&,*=)-S9S749745-.*=5*:9=U97*= 7*57S8*39S= )*= +&O43= 1.3S&.7*`= 1*8= &884(.&9.438= 2>(47-.?.*33*8= 6:.= 5*72*99*39= (*99*= -S9S749745-.*= 8439= ).;*78*8= *9= (477*8543)*39= A= )*8= -.894.7*8= S;41:9.;*8= ).++S7*39*8_= '.39S7U9= )*=(*= 8(-S2&= B.,_= ./=*89= )*= 84:1.,3*7= 1*8= 54.398=(422:38=A= 1S;41:9.43= )*=1&= 2>(4-S9S749745-.*=)&38=).++S7*398=S(48>89T2*8=*9=).++S7*39*8=97.':8=)47(-.)S*8`=2&.8=*3= &:(:3=(&8=.1=3*=8&,.9=):3*=S;41:9.43=1.3S&.7*=*9=4:=).7*(9.433*11*_= 1*99*= (&5&(.9S= A= 7*(*;4.7= ):= (&7'43*= )*8= (-&25.,3438= 2>(47-.?.*38= 5*:9= *=51.6:*7= 1&= +7S6:*39*= &55&7.9.43= )*= 2>(4-S9S749745-*8= 4:= )*= 2.=49745-*8$= (-*?= 1*8= 47(-.)S*8_= 1*8=(-&25.,3438= 2>(47-.?.*38= 8439=)*8= '&8.).42>(T9*8= 4:=)*8= &8(42>(T9*8`= )439= (*79&.38= +472*39= )*8= *(942>(47-.?*8_= 5&38= 1&884(.&9.43= *(942>(47-.?.*33*`= ,S3S7&1*2*39`= 1&:949745-*= +4:73.9= ):= (&7'43*= &:= (-&25.,343= 2&.8= 1*8= +1:== 8*2'1*39= &:88.=7S;*78.'1*8`=(422*=1*=24397*=1*8=(&8=)*=97&38+*798=)*=(&7'43*=):3*=*85T(*=)&7'7*=A= 1&:97*= ;.& =1*8=7S8*&:==2>(S1.*38=(422:38_=1*8=+1:==.3;*78S8=8439=*3(47*=5*:=(433:8=b==.18= 3*=8439=5&8=3S(S88&.7*2*39=5*72&3*398=a=.18=)S5*3)*39=)*=1&=8&.843`=)*=1&=5-S3414,.*=*9= )*8= (43).9.438= )S(1&.7*2*39= .2&7)= ]= 5:7&11`= ,**._= 5&38= 1*= (&)7*= )*8= 47(-.)S*8= 94:9*+4.8`= (*79&.3*8= 8439= 2>(4-S9S749745-*8= 94:9*= 1*:7= ;.*`= *9= 2&3.5:1*7&.*39= )43(= 1*:78= (-&25.,3438=2>(47-.?.*38=54:7=2&.39*3.7=(*8=+1:==.3;*78S8_=5&:97*8=2>(4-S9S749745-*8= 7*O4.;*39=&:88.=):=(&7'43*= ;.& =1*:78=(-&25.,3438=2>(47-.?.*38`=)*8=,142*742>(T9*8=6:.= +472*39=5&7=&.11*:78=)*8=*3)42>(47-.?*8=&;*(=)&:97*8=&:949745-*8_=5*=34:;*&:=(-*?=1*8= *3)42>(47-.?.*38=&884(.S8=A=)*8=&:949745-*8`=1*8=+1:==5*:;*39=U97*=7S;*78.'1*8`=(422*=1*= )S24397*39=1*8=+1:==)*=(&7'43*=*397*=51&39*8='*7&9`=,**-_= .= (*8= +1:== 7S;*78.'1*8= 8439= 5488.'1*8= )*= +&O43= 543(9:*11*= (-*?= 1*8= &:949745-*8`= .1= 8*7&.9=.39S7*88&39=)*=(4257*3)7*=(422*39=(*8=+1:==.3;*78S8=)*;.*33*39=5*72&3*398=(-*?= 1*8= 2>(4-S9S749745-*8`= *9= 6:*11*8= 2&3.5:1&9.438= 1*= 5*72*99*39= -47243*8`= ;.9&2.3*8`= 8.,3&:== *397*= 1*8= )*:== 5&79*3&.7*8c_= 5&:97*= 5&79`= 1*8= 47(-.)S*8= &55&79.*33*39= &:== 85&7&,&1*8= :3= 47)7*= )&3,.485*72*8_= .4:9*8= 1*8= 85&7&,&1*8= 8439= &884(.S*8= A= )*8= (-&25.,3438= *3)42>(47-.?.*38`= 4:= 3*= 8439= 5&8= 2>(47-.?S*8=a= 8*:1*8= 1*8= 47(-.)S*8= 8439= &884(.S*8=A=)*8='&8.).42>(T9*8_=1*99*=).++S7*3(*=.251.6:*=)*8=(-&3,*2*398=.25479&398=)*8= 2S(&3.82*8=)*=7*(433&.88&3(*=)*8=(-&25.,3438=2>(47-.?.*38`=5:.86:*11*=(477*8543)=A= :3=(-&3,*2*39=.25479&39=).)*39.9S=)*8=(-&25.,3438=2>(47-.?.*38=*9=):=+43(9.433*2*39= )*= 1&884(.&9.43= +1:== .3;*78S`= 1>8*= )*8= 5*149438_= 5*= 342'7*:8*8= 2>(4-S9S749745-*8= &55&79.*33*39= &:== '.1.&1*8`= &:== 5.48(47*&1*8= *9= &:== 85&7&,&1*8`= 974.8= 47)7*8= )*= 2434(49>1S)43*8= 2&.8= 8*:1*8= 1*8= 47(-.)S*8= 8439= &884(.S*8= A= )*8= (-&25.,3438= 2>(47-.?.*38= ).++S7*398= )*8= *85T(*8= ;*79*8_= 8S&324.38`= 1&= +7S6:*3(*= )*8= 2>(4-S9S749745-*8= )&38= (*8= 47)7*8= 8:,,T7*39= 6:*= (*79&.38= +&(9*:78= 5*:;*39= +&(.1.9*7= 1&55&7.9.43=)*=(*=24)*=)*=3:97.9.43`=6:.1=8&,.88*=)*=1&=574):(9.43=)*=5*9.9*8=,7&.3*8`=4:= )*=2S(&3.82*8=)*=2&3.5:1&9.43_='*=.89*3(*=)*=(*8=51&39*8=2>(4-S9S749745-*8=84:1T;*39=

98 )&:97*8= 6:*89.438= 8:7= 1*8= 7S8*&:== 2>(47-.?.*38= (422:38= *9= 1*:7= 7S,:1&9.43=a= *=.89*898.1= )*8= 2S(&3.82*8= )*= (43971*= )*8= +1:==)*= (&7'43*== 5*8= 8&3(9.438= 5*:;*398*11*8= U97*= .2548S*8= *3= (&8= )*= 8:77*=514.9&9.43= ):= 7S8*&:`= A= 1.389&7= )*= 1&= 8>2'.48*= 7-.?4'.:2 8 1S,:2.3*:8*8=94:,&&7)`=,***==!3=)&:97*8=9*72*8`=(*8=7S8*&:==2>(47-.?.*38=(422:38= 5*:;*398.18=(4397*88S1*(9.433*7=1*8=5:.98=)*=(&7'43*=9745=342'7*:==4:=9745=.25479&398== !9:).*7= 1*8= 5488.'1*8= (43971*8= )*= (*8= 7S8*&:== *9= )*= (*8= +1:== 5*72*997&.9= )*= 2.*:== (4257*3)7*= (*99*= 897&9S,.*= ,S3S7&1.89*= )*8= (-&25.,3438= 8>2'.49.6:*8`= 6:.= &55&7&9= S;41:9.;*2*39=89&'1*_= = 5&:97*=5&79`=94:9*8=(*8=2>(4-S9S749745-*8=(4143.8*39=)*8=-&'.9&98=8.2.1&.7*8`=)*8= 84:8= '4.8`= 9*25S7S8= 4:= 9745.(&:=_= 'S9:)*= )*8= &1'.348= 8:,,T7*= 6:*= 1&= 5*79*= )*= 1&= 5-4948>39-T8*= *89= (4397*88S1*(9.433S*= )&38= )*8= *3;.7433*2*398= 9745= 8*(8= *924:= 9745= 1:2.3*:=_= '&= 2>(4-S9S749745-.*= 3*89= )43(= 5488.'1*= 6:*3= 84:88'4.8= 842'7*`= *9= 5*72*9= )&.11*:78= )*= (4143.8*7= )*8= 2.1.*:== 41= 1*8= &:949745-*8= 8439= 1.2.9S8= 5&7= 1*= 2&36:*= )*= 1:2.T7*_=1*99*=(4397*88S1*(9.43=*3=2.1.*:=4:;*79=*89=827*2*39=&251.+.S*=5&7=1&=(425S9.9.43= &;*(=)*8=*85T(*8=&:949745-*8`=(425S9.9.43=6:.=*89=7*1B(-S*=*3=84:88'4.8_=1*99*=(425S9.9.43= 3&=5&8=S9S=57.8*=*3=(4259*=)&38=243=S9:)*`=2&.8=8*7&.9=A=S9:).*7`=54:7=(4257*3)7*=1*8= *=.,*3(*8=S(414,.6:*8=)*8=2>(4-S9S749745-*8`=4:=)*8=2.=49745-*8_= '&= (425S9.9.43= 5*:9= &:88.= U97*= *3;.8&,S*= *397*= 2.=49745-*8$= 4:= *397*= 2>(4-S9S749745-*8`= 6:.= 54:88*39= 84:;*39= &:= 2U2*= *3)74.9_= '*8= 2.=49745-*8$= 5*:;*39= +472*7=)*8=545:1&9.438=&88*?=)*38*8`=41=51:8.*:78=*85T(*8=8439=2S1&3,S*8=4'8*7;&9.438= 5*74833*11*8_= !=.89*898.1= :3*= (425S9.9.43= *397*= (*8= 2.=49745-*8$= )*85T(*8= ).++S7*39*8== %:='.*3=84398*11*8=&884(.S*8=A=)*8=(-&25.,3438=).++S7*398=*9=)*8=7S8*&:==).++S7*398=(422*= 1*= 8:,,T7*= 1*8= 5.5&(9.8 = *9= 1*8= *5-&1&39-*7&` = (*= 6:.= 1.2.9*7&.9= 1&= (425S9.9.43= *=(1:8.43= (425S9.9.;*=(43):.8&39=A=1&=+472&9.43=)*=,:.1)*8===03*=)*8=5*785*(9.;*8=)*=(*=97&;&.1= 8*7&.9=)*=2*8:7*7=1*=54:7(*39&,*=)-S9S749745-.*=)*=).++S7*39*8=*85T(*8=6:.=54:88*39=&:= 2U2*= *3)74.9= 4:= 343`= )&38= )*8= 545:1&9.438= )*= )*38.9S8= ).++S7*39*8`= *9= )S9:).*7= 1*:7= (479T,*= 2>(47-.?.*3_= .38.`= .1= 8*7&.9= 5488.'1*= )S9:).*7= (422*39= (*8= 2.=49745-*8$= *=514.9*39= 1*= 7S8*&:= 2>(47-.?.*3= (422:3`= 8.= 1->549-T8*= )*=(1:8.43= (425S9.9.;*= *89= ;&1&'1*=4:=343`=*9=8.=1&=)*38.9S=)*8=2.=49745-*8$=1.2.9*=4:=343=1*:7=)*,7S=)-S9S749745-.*_= 1*1&= 5*72*997&.9= &:88.= )S9:).*7= 1*= (429= )*8= 2.=49745-*8$= 54:7= 1*:78= (-&25.,3438= 2>(47-.?.*38_=03*=57*2.T7*=&5574(-*=54:77&.9=U97*=+43(9.433*11*=2*8:7*=)*=)*38.9S`=)*= 8.,3&9:7*=.84945.6:*`=)*=54:7(*39&,*=)*=2>(47-.?&9.43=*9=)S9*72.3&9.43=)*=1.)*39.9S=)*8= (-&25.,3438=2>(47-.?.*38=*9=54:77&.9=U97*=(4251S9S*=5&7=:3*=&5574(-*=5->14,S3S9.6:*`= *3=2*8:7&39=1&=).85*78.43=)*=1&=548.9.43=5->14,S3S9.6:*=)*=1*:78=5&79*3&.7*8=2>(47-.?.*38= )&38=1*8=5->14,S3.*8_=1*99*=&5574(-*=8*7&.9=&:88.=&551.(&'1*=&:==2>(4-S9S749745-*8=9*11*8= 6:*=1*8= 5->1147(-.8 _= = 143(*73&39= 1*8= 5->1147(-.8 `= .1= 7*89*= 343= 8*:1*2*39= A= S9:).*7= (*99*= 5488.'1*= (425S9.9.43= *397*= *85T(*8`= 5:.86:*= 51:8.*:78= *85T(*8= ) 5->1147(-.8 = 5*:;*39= (48*=.89*7`= (422*=A=54.=:9-*5= (+_ =&79.(1*= ==2&.8=&:88.=1*8=7*1&9.438=6:439=1*8= 5->1147(-.8 =&;*(=1&= (422:3&:9S=)*=(-&25.,3438=*(942>(47-.?.*38=a=1*:78=2>(47-.?*8=(477*8543)*398*11*8=A= 1*3;.7433*2*39= 14(&1= A= 574=.2.9S= )*8= 7&(.3*8= 4:= '.*3= (*8= 2>(47-.?*8= 84398*11*8= 8S1*(9.433S*8=5&72.=1*8=(-&25.,3438=).8543.'1*8=='->549-T8*=):3*=8S1*(9.;.9S=8:,,T7*= )*8= 2S(&3.82*8= )*= 7*(433&.88&3(*8= 2:19.51*8`= 9&3).8= 6:*= 1&'8*3(*= )*= 8S1*(9.;.9S= 84:1.,3*7&.9=1&'8*3(*=)*=2S(&3.82*=)*=7*(433&.88&3(*=*9=548*=)&:97*8=6:*89.438=8:7=1*8= )S+*38*8=;.88A8;.8=)*=(-&25.,3438=5&9-4,T3*8=4:=5&7&8.9*8_=5&:97*=5&79`=(*8=->549-T8*8=

99 8*7&.*39= A= 9*89*7= &:88.= 1478= )*= 1&= ,*72.3&9.43= )*8= 5->1147(-.8 `= &+.3= )*= 8&;4.7= 8.= 1&= 3438 85S(.+.(.9S=8&551.6:*=&:== ).++S7*398= 89&)*8= )*= )S;*1455*2*39_= .= 1&= ,*72.3&9.43=*89= 343= 85S(.+.6:*`= &1478= 1*8= 5->1147(-.8 = 5*:;*39= ).85*78*7= A= ).89&3(*= *9= )&38= )*8= 2.1.*:== 97T8= ).;*78`=2&.8=8.=1&=,*72.3&9.43=*89=85S(.+.6:*=4:=51:8=8S1*(9.;*`=(*(.=(4397.':*7&.9=A=2*997*= *3= S;.)*3(*= :3*= 85S(.&1.8&9.43= +43(9.433*11*= *9= 7*/4.3)7&.9= 1->549-T8*= )*= 8S1*(9.;.9S_= $S84:)7*= (*8= 6:*89.438= 5*72*997&.9= )*= 2.*:== (4257*3)7*= 1&= 7&7*9S= )*= (*8= *85T(*8= 2>(4-S9S749745-*8_= = :8)*1A= )*8= 7S+1*=.438= 8:7= 1&= 85S(.+.(.9S`= (*8= 6:*89.438= 5*72*99*39= )*= 2.*:== (433&97*= 1*8= *=.,*3(*8= S(414,.6:*8= )*85T(*8= 7&7*8`= )439= 1&= 8:7;.*= )S5*3)= )*= 7S8*&:== 2>(47-.?.*38= *9= 5&7+4.8= )&7'7*8= ;4.8.38_= 5*= 9*11*8= S9:)*8= 5*72*99*39= )&2S1.47*7= 1&= (438*7;&9.43=)*8=47(-.)S*8=2>(4-S9S749745-*8=349&22*39= (+ _=&79.(1*=`=&79.(1*= _=5&38= 1&=2*8:7*=41=(*8=47(-.)S*8=8439=&884(.S*8=A=)*8=7S8*&:==2>(S1.*38=(422:38=*9=)*8=&7'7*8`= 5*79:7'*7=(*8=7S8*&:==)43(=1*=841=4:=(4:5*7=1*8=84:7(*8=)*=(&7'43*=1*8=&7'7*8=7*;.*39=A= 9:*7=1&=545:1&9.43=)47(-.)S*=2>(4-S9S749745-*_=5*=2U2*`=1&=)S(4:;*79*=)*=342'7*:8*8= *85T(*8= 2.=49745-*8`= *9= 1&= 549*39.&1.9S= )&:97*8= *85T(*8= A= U97*= 2.=49745-*8`= 2U2*= *3= )*-478= )*8= 47(-.)S*8`= 84:1.,3*= 1&= 3S(S88.9S= )*= 57*3)7*= *3= (4259*= 1*8= &884(.&9.438= 2>(47-.?.*33*8=)&38=1*8=574/*98=)*=(438*7;&9.43=)*8=*85T(*8=;*79*8= (+_ &79.(1*= =b=Z*991*7= *9= &1_` = ,**-=b= *1488*= *9= &1_` = ,**.&_= '&= (411&'47&9.43= &;*(= 1*= /&7).3= '49&3.6:*= )*= 1&= $*.3*= .7.0.9=*3=.-&1&3)*= (+ _= &79.(1*=`= =&=5*72.8=).3.9.*7=)*8=574/*98=).)*39.+.(&9.43=)*8= 5&79*3&.7*8= 2>(47-.?.*38= )47(-.)S*8= (-14745->11.*33*8= 7&7*8`= &+.3= )&2S1.47*7= 1*:7= (43).9.438=)*=(:19:7*=&:=/&7).3='49&3.6:*=*9=)S;.9*7=1*8=97&3851&39&9.438=.3:9.1*8_='S6:.5*= )*=&&2&94=&=5&7=*=*251*=7S:88.=A=(:19.;*7= *5-&1&39-*7&=+&1(&9&= &&2&94=]=<&8*`=,**2=)*= 1&= ,7&.3*= A= 1&= 51&39*= +1*:7.*= 5*3)&39= /= &38`= &1478= 6:*= (*99*= *85T(*= *89= 5488.'1*2*39= 2.=49745-*_= 5*= 2U2*`= (*99*= S6:.5*= &= (:19.;S= )*8 = 5.54,.:2= 748*:2 = *3= 2.(74(482*= &&2&94=*9=&1_`=,**/=*9=4'9*3:=1&=+147&.843`=*3=(:19.;&39= (43/4.39*2*39= 1&= 51&39*= *9= 843= (-&25.,343= 2>(47-.?.*3_= '&= 5488.'.1.9S= )*= (:19:7*= *3= 2.(74(482*= 5*:9= 343= 8*:1*2*39= 5*72*997*= 1&= 7S.3974):(9.43= )*= (*79&.3*8= *85T(*8= .3= 8.9: = 4:= 1*:7= (438*7;&9.43= )&38= )*8= /&7).38= '49&3.6:*`= 2&.8= 5*:9= &:88.= 4:;7.7= 1&= ;4.*= A= )*= 34:;*11*8= *=5S7.2*39&9.438= 54:7= S9:).*7=1.39*7&(9.43=)*8=2>(4-S9S749745-*8=&;*(=1*:78=(-&25.,3438=2>(47-.?.*38_=5*=9*18= 2.(74(482*8=5*72*997&.*39=5&7=*=*251*=)*=2*8:7*7=1*8=(:98=)*=(*99*=&884(.&9.43=54:7=1*8= )*:== 5&79*3&.7*8_= 1*8= 2.(74(482*8= 8*7&.*39= :9.1*8= 54:7= ).88S6:*7= 51:8= +.3*2*39= 1&= )>3&2.6:*= )*8= +1:== )*= (&7'43*`= 349&22*39= (-*?= 1*8= *85T(*8= 2.=49745-*8`= &+.3= )*= ).89.3,:*7= 1*++*9= )*= 1&= 5-S3414,.*`= )*= 1S(1&.7*2*39`= 4:= ):= 8.9*= 8:7= 1*8= ;&7.&9.438= )*= 1&= 8.,3&9:7*=.84945.6:*=*3= +- 1_= = 5&38= 1&= 2*8:7*= 41= 1*8= 47(-.)S*8= 439= :3*= ;&1*:7= 8>2'41.6:*= .25479&39*= 54:7= )*= 342'7*:== '49&3.89*8= 4:= &2&9*:78`= *9= :3*= ;&1*:7= (422*7(.&1*= 94:9= &:88.= S1*;S*`= 1*8= 57S1T;*2*398= )&38= 1&= 3&9:7*= 8439= +7S6:*398= *9= )S;&89&9*:78_= '&= ).++:8.43= )*8= (433&.88&3(*8= 8:7= 1*:7= S(414,.*= *89= )43(= *88*39.*11*`= ):3*= 5&79= 54:7= 7S):.7*= (*8= 57S1T;*2*398`=*9=)&:97*=5&79=54:7=24397*7=1.3:9.1.9S=)*=(*8=57S1T;*2*398=)&38=1&=2*8:7*= 41= 1*8= 2>(4-S9S749745-*8= *9= 1*8= 2.=49745-*8$= 3*= 5*:;*39= 8:7;.;7*= A= 1&= 97&3851&39&9.43= &)4;80>`=+30/_=03*=)*8=5*785*(9.;*8=.25479&39*8=)*=2&=9-T8*=8*7&=)43(=)*=(439.3:*7=A= ;:1,&7.8*7= (*8= (433&.88&3(*8= 34:;*11*8= 8:7= 1&= 2>(4-S9S749745-.*= *9= 1&= 2.=49745-.*= (+_ = &79.(1*= `= *9= 8*38.'.1.8*7= A= 1.25479&3(*= )*= 1&= 57.8*= *3= (4259*= )*8= .39*7&(9.438= 2>(47-.?.*33*8=54:7=1&=(438*7;&9.43=)*8=47(-.)S*8_= =

100 = =

101 ;; =

&3*3`=5_=A_`=._=F_=A:>5*7`=._=4*0-4:9`=&3)=$_=B_= 4*0897&_=,***&_=->14,*3*9.(=7*1&9.438-.58=.3= 9-*=,*3:8= **'*142& ='&8*)=43=.+=&3)=,=8*6:*3(*8`=<.9-=85*(.&1=*25-&8.8=43=9-*= **'*142&= (7:89:1.3.+472*= (4251*=_=#>(414,.&= 3, a,038,2+_=

'&).*`=<_=1_`=0_=:998*55`= _= *'&:*7`=_=B&((.4`=_=43+&39*`=&3)=#_=_=*1488*_=,**0_= *5-&1&39-*7&= 143,.+41.& =8*499.*&*`=%7(-.)&(*&*=.8=2.=49745-.(a=&=(425&7&9.;*=89:)>='*9<**3=,7**3=&3)= 3435-4948>39-*9.(= .3).;.):&18_= 1&3&).&3= <4:73&1= 4+= 49&3>8$*;:*= 1&3&).*33*= 5*= 49&3.6:*= 2. a+.0,8+.11_=

1*=&3)*7`= _= <_`= &3)= _= 4,'*7,_= +320_= !(942>(477-.?&8= 4+= .745.(&1= 3,.485*724:8= .7**8_= 8*<= ->9414,.89= +*, a/.+8/.3_=

1*=&3)*7`=_=<_`=&3)=_=_='**_=,**/_=#>(477-.?&8= &3)=*(48>89*2=574(*88*8=.3=9745.(&1=7&.3=+47*89a= .251.(&9.438=+47=).;*78.9>_=&,*8=+0/8,*-= .3 =.49.(=39*7&(9.438=.3=9-*=.745.(8a=.-*.7=$41*=.3= 9-*=#&.39*3&3(*=4+=5*(.*8=5.;*78.9>_=1&2'7.),*=03.;*78.9>=7*88_=

29-47`=<_=_=+33/_= .,-*7=1&39=$*85.7&9.43=&3)=98=$*1&9.438-.58=94=-4948>39-*8.8_=&,*8=1+8+*+= .3 =!_85_=(-:1?*=&3)=_=1&1)<*11`=*).9478_=!(45->8.414,>=4+=-4948>39-*8.8_=57.3,*7`=8*<= &470_=

7(-.'&1)`=<_=#_=,**3_=.-*=:??1*=4+=1&89.)=!;41:9.43_=1:77*39=.414,>= +3 a$2+8$22_=

7).99.`=<_`=&3)=_=A_=_= -&3._=,***_=.&381*>=7*;.*<=84_=++*=8=8:2*7.(&1=&3)=5->8.(&1=5745*79.*8=4+= 47(-.)=8**)8=&3)=9-*.7='.414,.(&1=.251.(&9.438_=8*<=->9414,.89= +./ a-018.,+_=

7341)`= _= !_`= &3)= B_= ':9?43._= ,**1_= 5.;*78.9>= &3)= -489= 7&3,*= 4+= +41.&7= +:3,&1= *3)45->9*8a= 7*= 9745.(&1=1*&;*8='.4).;*78.9>=-4985498=!(414,>=22a/.+8/.3_=

7341)`=_=!_`=Z_=#&>3&7)`= _=_= .1'*79`=_=5_=141*>`=&3)=._=_=A:78&7_=,***_=7*=9745.(&1=+:3,&1= *3)45->9*8=->5*7).;*78*=!(414,>='*99*78=-a,018,1._=

&)*(0`= B_= F_`= _= .(-*70*?`= _= 84,:*8`= 1_= .*1`= &3)= <_= -&8-,-&.*_= ,**/_= 48985-494= 8>39-*9.(= +7&(9.43&9.43=4+=89&'1*=(&7'43=.84945*8='*9<**3=51&39=47,&38=8=&=<.)*857*&)=5-*342*343_= $&5.)=1422:3.(&9.438=.3=#&88=5*(9742*97>= +3 a+-2+8+-3+_=

&702&3`=._=<_`=<_=$_=#(8*&1`=_= _='.2`= _=14&9`= _=_=17442`=8_=5_=&4:3,`=&3)=1_=F_=)*&25-.1.8_= ,**1_= #.94(-43)7.&1= 58= 8:,,*898= &9= 1*&89= ++= 47.,.38= 4+= 5&7&8.9.82= .3= &3,.485*728= &3)= 7*;*&18=,*342.(=(-.2*7.82=.3=5&7&8.9.(=51&398_=2(=!;41:9.43&7>=.414,>= 1a8_=

&77*99`=1_=B_`=&3)=<_=C_=B7*:)*389*.3_=,**2_=#41*(:1&7=*;41:9.43=4+=7'('=.3=9-*=2>(4-*9*749745-.(= (47&17449= 47(-.)8= 47&1147-.?& = &,3*'.3`= %7(-.)&(*&*_= #41*(:1&7= ->14,*3*9.(8= &3)= !;41:9.43= .1 a00/8013_=

&9*2&3`=$_=#_`=_=#_= 411.3,8<479-`=<_=6:.7*11`=&3)=#_='_= 411.3,8<479-_=,**/_=->14,*3*9.(8a= 8*499.*&*_= .3 =7.),*43_=_#_`=_=<_=17.''`=#_=F_=1-&8*`=&3)=B_=8_=$&82:88*3`=*).9478_= *3*7&=%7(-.)&(*&7:2=.a=!5.)*3)74.)*&*_=%=+47)=03.;*78.9>=7*88`=%=+47)`=0A_=

*73&7)`=8_=+3*3_='S;41:9.43=)&38=1&=8>2'.48*=)*8=47(-.)S*8=*9=1*:78=(-&25.,3438=(422*38&:=_= 33&1*8=)*8=(.*3(*8=8&9:7*11*8=)*=&7.8= 3a+8+30_=

102 *;*7`=<_=5_=,**,_= 489885*(.+.(.9>=4+=#=+:3,&1=545:1&9.43=,74<9-=7&9*8=(&3=,*3*7&9*=+**)'&(0=43= 51&39=,74<9-_=1&39=&3)=4.1= ,.. a,2+8,3*_=

*>71*`= _=B_`=&3)=_=!_=2.9-_=+33-_=!=(*88.;*=(&7'43=57*;*398=,7**3.3,=4+=1*&;*8=.3=2>(477-.?&1= 8**)1.3,8=4+=9-*=9*77*897.&1=47(-.)= 77(-.8=247.4 _='.3)1*>&3&= 2a31833_=

*>71*`= _=B_`=_=!_=2.9-`=$_='_=*9*7843`=&3)=1_=#_=#_=B7&3(4_=+33/_=14143.?&9.43=4+=%7(-.88#47.4= 7494(4728= '>= &= #>(477-.?&1= B:3,:8= 8= !++*(98= 4+= 8.974,*3= 8:97.9.43= &3)= 1>5-48&9*= .3= #4).+>.3,= 9-*= $*85438*8_= 1&3&).&3= <4:73&1= 4+= 49&3>8$*;:*= 1&3&).*33*= 5*= 49&3.6:*= 1- a++,28++.*_=

.)&7943)4`=#_=_=,**/_=.-*=*;41:9.43&7>=*(414,>=4+=2>(48-*9*749745->_=8*<=->9414,.89= +01 a--/8 -/,_=

.)&7943)4`= #_= _`= &3)= ._= 5_= 7:38_= ,**+_= !=97*2*= 85*(.+.(.9>= .3= *5.5&7&8.9.(= #43497454.)*&*= !7.(&(*&*a= <.)*857*&)= 5->14,*3*9.(= &3)= ,*4,7&5-.(&1= 897:(9:7*_= #41*(:1&7= !(414,>= +* a,,2/8,,3/_=

.)&7943)4`=#_=_`=&3)=._=5_=7:38_=,**,_=B.3*81*;*1=2>(477-.?&1=85*(.+.(.9>=.3=9-*=#43497454.)*&*= !7.(&(*&*a=85*(.+.(.9>=+47=+:3,&1=85*(.*8=,74:58_=#41*(:1&7=!(414,>= ++ a//18/03_=

.)&7943)4`=#_=_`=&3)=._=5_=7:38_=,**/_=%3=9-*=47.,.38=4+=*=97*2*=2>(477-.?&1=85*(.+.(.9>=.3=9-*= #43497454.)*&*=!7.(&(*&*a=5*7+472&3(*=97&)*84++8=):7.3,=8**)=,*72.3&9.43=&3)=8**)1.3,= )*;*1452*39_=#41*(:1&7=!(414,>= +. a+/.38+/0*_=

.)&7943)4`=#_=_`=_=:7,-&7)9`= _= *'&:*7`=._=5_=7:38`=&3)=5_=<_=$*&)_=,**._=1-&3,.3,=5&793*78=.3= 9-*= )&70a= .84945.(= &3)= 241*(:1&7= *;.)*3(*= 4+= *(942>(477-.?&1= 1.&.8438= '*9<**3= +47*89= 47(-.)8=&3)=97**8_=74(**).3,8=4+=9-*=$4>&1=4(.*9>=4+='43)43=*7.*8=8.414,.(&1=(.*3(*8= ,1+ a+1338+2*0_=

.)&7943)4`=#_=_`= _=!0`= _=F&11&3)*7`=&3)=_=4)*789742_=,**+_=54=3:97.*39=&)).9.438=&19*7=(&7'43= = 8.30=897*3,9-=4+=*(942>(477-.?&1=+:3,.=8*<=->9414,.89=+/+a/.-8//*_= = .)&7943)4`=#_=_`=_=#_=A7*9?*7`=!_=#_=.3*`=&3)=._=5_=7:38_=,***_= .,-=7449=(43(*397&9.43=&3)= :3*;*3= *(942>(477-.?&1= ).;*78.9>= 3*&7= #&7(4)*8= 8&3,:.3*&= !7.(&(*&*a= = (-*&9*7= 9-&9= 89.2:1&9*8=.98=;.(9.28=2*7.(&3=<4:73&1=4+=49&3>=21 a+12-8+122_=

.)&7943)4`=#_=_`=&3)=5_=<_=$*&)_=,**2_=B:3,&1=85*(.+.(.9>='4991*3*(08=):7.3,=47(-.)=,*72.3&9.43= &3)=)*;*1452*39_=#41*(:1&7=!(414,>= +1 a-1*18-1+0_=

.)&7943)4`=#_=_`=5_=$*)*(0*7`=_= ./7.`=_=F.*20*3`=._=5_=7:38`='_=542.3,:*?`=_=*78.(`=<_=$_= '*&0*`= &3)= 5_= <_= $*&)_= ,**,_= !5.5&7&8.9.(= 51&398= 85*(.&1.?*)= 43= &7':8(:1&7= 2>(477-.?&1= +:3,._=8&9:7*= .+3 a-238-3,_=

.1,*7`= _= F_`= 0_= (-7*.'*7`= &3)= %_= '_= '&3,*_= +32._= 5*9*72.3&9.43= 4+= '*&+= *&98$*8.89&3(*= 8= 1425&7&9.;*= 3;*89.,&9.43= 4+= 1-14745->11= B1:47*8(*3(*= 1-&3,*8= &3)= ..88:*= 8*(748.8= #*9-4)8_=%*(414,.&= 0- a,/08,0,_=

/702&3`=!_=+30*_= 4349745&=*>545.9>8 ='_=;=&3=!5.5&7&8.9*=43=.7**=$4498_=->8.414,.&=1&39&7:2= == +- a-*2=8=-,1_=

103 /702&3`=%_`=&3)=_=5*22.38)&28_=+33/_=$*,:1&9.43=4+=-4948>39-*8.8='.,-9=!3*7,>=1&59:7*`= 143;*78.43`=&3)=5.88.5&9.43=.3='*&;*8=4+= .,-*7=1&398_=&,*8=+18.1= .3 =!_85_=(-:1?*=&3)=_= 1&1)<*11`=*).9478_=!(45->8.414,>=4+=-4948>39-*8.8_=57.3,*7`=8*<=&470_=

4:73S7.&8`=#_`=5_=7&99`=&3)=1411*(9.+=)*=1&=4(.S9S=B7&3O&.8*=)%7(-.)45-.1.*_=,**/_='*8=%7(-.)S*8= )*=B7&3(*`=*1,.6:*=*9=':=*2'4:7,`=,3)=*).9.43_=.4945*`=#T?*_=

7:,341.`=!_`=A_=._= :'.(0`=_=C43(&*22*7*7`=_=1_= F43,`= &3)= _=5_=B&76:-&7_=+322_=1477*1&9.43= '*9<**3=9-*=1&7'43=84945*=5.8(7.2.3&9.43=.3='*&+=9&7(-=&3)=:,&78=4+=18-=1&398=&3)=9-*= $&9.4= 4+= 39*7(*11:1&7= &3)= 92485-*7.(= &79.&1= 7*88:7*8= 4+= 1&7'4385.4=.)*_= 1&39= 5->8.414,>= 22 a+.+28+.,._=

7:3)7*99`= #_= 1_= ,**,_= 14*;41:9.43= 4+= 74498= &3)= 2>(477-.?&8= 4+= 1&3)= 51&398_= 8*<= ->9414,.89= +/. a,1/8-*._=

7:38`= ._= 5_= +33/_= .-4:,-98= 43= 9-*= 74(*88*8= .-&9= #&.39&.3= '4(&1= 5*(.*885.;*78.9>= 4+= !(942>(477-.?&1=B:3,._=1&39=&3)=4.1= +1* a0-81-_=

7:38`= ._= 5_`= #_= _= .)&7943)4`= &3)= 5_= '_= .&>147_= ,**,_= 489= 85*(.+.(.9>= .3= *(942>(477-.?&1= (422:3.9.*8a=F-&9=)4=9-*=*=(*59.438=9*11=:8=39*,7&9.;*=&3)=1425&7&9.;*=.414,>= ., a-/,8 -/3_=

7:38`=._=5_`=&3)=5_=<_=$*&)_=,***_=3=;.974=,*72.3&9.43=4+=3435-4948>39-*9.(`=2>(48-*9*749745-.(= = 51&398=89.2:1&9*)='>=+:3,.=.841&9*)=+742=9-*=&):19=51&398_=8*<=->9414,.89=+.2a--/8-.,_= = 7?4804`=!_=,**,_=.-*=)>3&2.(8=4+= .89*7&=4;&9& =545:1&9.438=43=2.3*7&1=.81&3)8=.3=9-*=.*'7?&= 3&9.43&1== 5&70_=(9&=4(.*9&9.8=49&3.(47:2=4143.&*= 1+ a,.-8,/+_= = :7,*++`= _=+3-,_=&5745->9.82=:3)=>2'.48*_= :89&;=B.8(-3*7`=<*3&a+/.8+/3_=

:7,*++`= _=+3-0_=&2*30*.2:3,=)*8=%7(-.)**3_= :89&;=B.8(-3*7`=<*3&a+.+8+./_=

:7,*++`= _=+3/3_=#>(477-.?&=4+=47(-.)8_=&,*8=-0+8-3/= .3 =1_='_=F-.93*7`=*).947_=.-*=47(-.)8_=$43&1)= 7*88`=8*<=&470_=

1&2*743`=5_=5_`=_=<4-3843`=5_=<_=$*&)`=&3)=<_=$_='*&0*_=,**2_= .;.3,=&3)=7*(*.;.3,a=2*&8:7.3,=9-*= (&7'43=(489=4+=2>(477-.?&8=.3=9-*=,7**3=47(-.)`= .44)>*7&=7*5*38 _=8*<=->9414,.89= +2* a+108 +2._=

1&2*743`=5_=5_`=<_=$_='*&0*`=&3)=5_=<_=$*&)_=,**0_=#:9:&1.89.(=2>(477-.?&=.3=47(-.)8a=*;.)*3(*=+742= 51&398+:3,:8=(&7'43=&3)=3.974,*3=97&38+*78=.3=9-*=,7**381*&;*)=9*77*897.&1=47(-.)= .44)>*7&= 7*5*38 _=8*<=->9414,.89= +1+ a.*/8.+0_=

1&2*743`= A_= #_= ,**._= 09.1.9>= 4+= 51&89.)= 58& = ,*3*= 8*6:*3(*8= +47= .3;*89.,&9.3,= .397&+&2.1.&1= 7*1&9.438-.58=<.9-.3=%7(-.)&(*&*_=#41*(:1&7=->14,*3*9.(8=&3)=!;41:9.43= -+ a++/18++2*_=

1&25'*11`=!_=%_=+30._=.-*=+:3,&1=&884(.&9.43=.3=&=(4143>=4+= .&8974).&=8*8&24.)*8 _=.7&38&(9.438=4+=9-*= $4>&1=4(.*9>=4+=8*<=Z*&1&3)=.414,.(&1=(.*3(*8=,a,-18,.0_=

1&25'*11`=!_=%_=+31*_=.-*=+:3,&1=&884(.&9.43=4+= 24&3.&=&:897&1.8 _=.7&38&(9.438=4+=9-*=$4>&1=4(.*9>=4+= 8*<=Z*&1&3)=.414,.(&1=(.*3(*8= +, a/8+,_=

104 1-&8*`= #_= F_`= A_= #_= 1&2*743`= $_= '_= &77*99`= &3)= <_= C_= B7*:)*389*.3_= ,**-_= 58= )&9&= &3)= %7(-.)&(*&*=8>89*2&9.(8a=&=3*<=5->14,*3*9.(=(1&88.+.(&9.43_=&,*8=03823= .3 =A_=#_=5.=43`=_= _= A*11`= $_= '_= &77*99`= &3)= _= <_= 17.''`= *).9478_= %7(-.)= (438*7;&9.43_= 8&9:7&1= .8947>= :'1.(&9.438`=A49&=A.3&'&1:`=&'&-`=#&1&>8.&_=

11&>`= A_`= &3)= _= 6_= A4;*7_= +330_= .-*= $*)= I:**3= >549-*8.8= &3)= 51&3925&9-4,*3= .39*7&(9.438_= 33:&1=$*;.*<=4+=->945&9-414,>= -. a,38/*_=

11*2*398`=#_=_=+322_=%7(-.)=2>(477-.?&1=&884(.&9.438_='.3)1*>&3&= -a1-820_=

17.''`= _= <_`= _= _= A*11`= A_= F_= 5.=43`= &3)= $_= '_= &77*99_= ,**-_= %7(-.)= (438*7;&9.43a= &= ,14'&1= 5*785*(9.;*_= &,*8= +8,/= .3 =A_=F_=5.=43`=_=_=A*11`=$_='_=&77*99`=&3)=_=<_= 17.''`= *).9478_= %7(-.)= (438*7;&9.43_= 8&9:7&1= .8947>= 5:'1.(&9.438= 473*4`= A49&= A.3&'&1:`= &'&-`= #&1&>8.&_=

1:11.3,8`=A_=F_`=._=#_=?&74`=&3)=._=5_=7:38_=+330_=!;41:9.43=4+=*=97*2*=85*(.&1.?&9.43=<.9-.3=&= 1.3*&,*=4+=*(942>(477-.?&1=*5.5&7&8.9*8_=8&9:7*= -13 a0-800_=

1:79.8`= <_= ._= +3-1_= 843885*(.+.(.9>= 4+= 47(-.)= 2>(477-.?&1= +:3,._= 74(**).3,8= 4+= 9-*= 4(.*9>= +47= !=5*7.2*39&1=.414,>=&3)=#*).(.3*= -0 a.-8.._=

1:79.8`= <_= ._= +3-3_= .-*= 7*1&9.43= 4+= 85*(.+.(.9>= 4+= 47(-.)= 2>(477-.?&1= +:3,.= 94= 9-*= 574'1*2= 4+= 8>2'.48.8_=2*7.(&3=<4:73&1=4+=49&3>= ,0 a-3*8-33_=

5&-1'*7,`= _= ,**+_= 1422:3.9>= *(414,>= 4+= *(942>(477-.?&1= +:3,.a= &3= &);&3(.3,= .39*7).8(.51.3&7>= +.*1)_=8*<=->9414,.89= +/* a///8/0,_=

5&2*8.3`=1_`=&3)=1_='*1&7,*_=,**-_=1&7'43=.84945*=(42548.9.43=4+=(:77*398>*&7=8-4498=+742= &&,:8= 8>1;&9.(& =.3=7*1&9.43=94=,74<9-`=7*85.7&9.43=&3)=:8*=4+=7*8*7;*8_=1&39=1*11=&3)=!3;.7432*39= ,0 a,*18,+3_=

5&2*8.3`=1_`=_=$&2'&1`=&3)=$_=<4++7*_=+332_=*&843&1=&3)=&33:&1=(-&3,*8=.3=1*&+=)*19&=18+-=.3=9<4= (484((:77.3,= #*).9*77&3*&3= 4&08a= 7*1&9.438= 94= 1*&+= ,74<9-= &3)= )74:,-9= 574,7*88.43_= B:3(9.43&1=!(414,>= +, a112812/_=

5&7<.3`=1_=$_=+20,_=%3=9-*=;&7.4:8=(4397.;&3(*8='>=<-.(-=7.9.8-=&3)=+47*.,3=47(-.)8=&7*=+*79.1.8*)= '>=.38*(98`=&3)=43=9-*=,44)=*++*(98=4+=.39*7(7488.3,_=<4-3=#:77&>`='43)43_=

5&<843`=._=!_`=_=#&2'*11.`=_= _=1&2'4*(0`=_= _=.*251*7`=&3)=A_=_=.:_=,**,_=9&'1*=.84945*8=.3= 51&39=*(414,>_=33:&1=$*;.*<=4+=!(414,>=&3)=>89*2&9.(8= -- a/*18//3_=

5*&73&1*>`=<_=5_=F_=,**1_=B:79-*7=&);&3(*8=.3=47(-.)=2>(477-.?&1=7*8*&7(-_=#>(477-.?&= +1 a.1/8.20_=

5*&73&1*>`=<_=5_=F_`=&3)=_=B_='*=74(6:*_=,**0_=#41*(:1&7=.)*39.+(&9.43=4+=9-*=57.2&7>=7449=+:3,&1= *3)45->9*8=4+= ,.54).:2=-&2.1943.&3:2 =%7(-.)&(*&*_=:897&1.&3=<4:73&1=4+=49&3>= /. a.218 .3+_=

5*1+47,*`= _= +332_= 5*8= 5.5&(9.8= -*11*'47.3* = '_= 1$8.Z= )S54:7;:8= )*= (-14745->11*= )&38= 1*8= *3;.7438=)*=7:=*11*8_=8&9:7&1.89*8=*1,*8= 13 a+,.8+-*_=

5*11`= _`= $_= &32**`= _= ':2>43,`= &3)= _= ':2>43,_= ,**/_= !(942>(477-.?&1= +:3,.= .3= )7>= &3)= <*9= ).59*74(&75= +47*898= .3= 3479-*73= .-&.1&3)= 8= ).;*78.9>= &3)= :8*= &8= +44)_= .3 = 8.&= &(.+.(= 884(.&9.43=4+=B47*897>=$*8*&7(-=389.9:9.438`=29-=<4708-45_=

105 5*22.,8&)&28`= _`= &3)= F_= F_= )&28_= +33,_= -4945749*(9.43= &3)= %9-*7= $*85438*8= 4+= 1&398= 94= .,-='.,-9=97*88_=33:&1=$*;.*<=4+=1&39=->8.414,>=&3)=1&39=#41*(:1&7=.414,>= .- a/338 0,0_=

)*3=&00*7`= _=1_`= _=1_=Z:((&7*114`=._=F_=A:>5*7`=&3)=#_=!_=8447)*1448_=,**._=!;41:9.43=&3)=-489= 85*(.+.(.9>=.3=9-*=*(942>(477-.?&1=,*3:8= *((.3:2 _=8*<=->9414,.89= +0- a,*+8,+/_=

5*82&7&.8`= !_`= _= '&33*1:(`= &3)= <_= '&,3*1_= +332_= 5.7*(9= &251.+.(&9.43= 4+= 1*3,9-= 541>2475-.828= 5'`= 47= -4<= 94= ,*9= &3)= (-&7&(9*7.?*= 3*<= ,*3*9.(= 2&70*78= .3= 2&3>= 85*(.*8_= 8:(1*.(= (.)8=$*8*&7(-= ,0 a+./28+.0/_=

5.(0.*`= _= _`= $_= 1_= :?&`= _= !_= A7&?*<80.`= &3)= _= _= $*.(-_= ,**._= -&7*)= *(942>(477-.?&1= +:3,.= '*9<**3=&=-*7'&(*4:8=5*7*33.&1= **1.&39-*2:2='.(03*11.. =&3)=4&0= 9:*7(:8 =8**)1.3,8_=8*<= ->9414,.89= +0. a-1/8-2,_=

54'?-&380>`= ._= +31-_= 849-.3,= .3= .414,>= #&0*8= *38*= *=(*59= .3= '.,-9= 4+= !;41:9.43_= 2*7.(&3= .414,>=.*&(-*7= -/ a+,/8+,3_=

542.3,:*?`= '_= _`= &3)= _= *78.(_= ,**._= .-*= 84:9-*732489= 2>(48-*9*749745-.(= 51&39`= 7&(-3.9.8= :3.+147& a=7449=2475-414,>=&3)=&3&942>_=#>(414,.&= 30 a++.-8++/+_=

5472439`= '_`= $_= 5*11*8C*)4;*`= <_= #_= *88.*7*`= #_= 488&*798#(A*>`= &3)= _= (-&9?_= ,**3_= = 34;*1= 2*(-&3.82= 2&.39&.3.3,= +147&1= 541>2475-.82a= <-.9*8+14<*7*)= 2475-8= &8= -*15*78= .3(7*&8.3,=7*574):(9.;*=8:((*88=4+=5:751*=2475-8=.3=&=+44)8)*(*59.;*=47(-.)_=8:'2.99*)=94= 8*<=->9414,.89_=

5473.*7`= _`= B_= #:34?`= &3)= _= %_= 1-*594:_= ,**2_= 11**= !++*(9= &3)= *1+8B*79.1.?&9.43= 3-*72&5-74).9*8a= $*574):(9.;*= 88:7&3(*= .3= &= 97:(9:7*)= #*9&545:1&9.43_= !;41:9.43= 0, a,//28,/03_=

54880*>`=#_= _`=$_= _='.3)*72&3`=&3)='_=4*782&_=+33*_=1&7'438.30=9.2:1&9.43=4+=-4948>39-*8.8= .3=54:,1&88B.7=**)1.3,8='>=42*=!(942>(477-.?&8_=8*<=->9414,.89= ++/ a,038,1._=

54:-&3`= _= F_`= &3)= 5_= #_= $.??4_= ,**/_= ->14,*3*9.(= ).;*7,*3(*= .3= &= 14(&1= 545:1&9.43= 4+= 9-*= *(942>(477-.?&1=+:3,:8= *34(4((:2=,*45-.1:2 _=8*<=->9414,.89= +00 a,0-8,1+_=

57*881*7`=$_='_=+33*_=.-*=8*499.*&*=.3=47(-.)=(1&88.+.(&9.43_='.3)1*>&3&= /a+*,8+*3_=

B*1)2&33`=A_=_=+33+_=.858=38*79.43=#:9&,*3*8.8=.3= 7&'.)458.8 =8=#:9&9.43&1=5*(97:2_=1&39= <4:73&1= +a1+82,_=

B7&30*`=._`='_=**30*3`=#_=547.3,`=_=A4(>&3`=&3)=$_=,*7*7_=,**0_=7':8(:1&7=2>(477-.?&1=+:3,.=4+= 9-*= 142:88,74:5= = 1.3*&,*= 142*7&1*8b= 142*742>(49&= )*9*(9*)= .3= 2>(48= -*9*749745-.(=51&398=+742=9745.(&1=+7.(&_=#>(414,.(&1=74,7*88= /a,.8-+_=

B:72&3`=._=!_`=&3)=<_=#_=.7&55*_=+31+_=->14,*3>=&3)= !(414,>= 4+= #>(49745-.(= (-14745->114:8= 3,.485*728_=I:&79*71>=$*;.*<=4+=.414,>= .0 a,+38]_=

&3,*`= _= 1_`= &3)= _= #_= F*89_= +33._= 39*7&(9.438= '*9<**3= 7':8(:1&7= #>(477-.?&1= B:3,.= &3)= B41.&78B**).3,=38*(98=.3= 1&39&,4=1&3(*41&9& ='_=8*<=->9414,.89= +,2 a13821_=

&7)*8`= #_= ,**,_= 3= 47(-.)8+:3,:8= 2&77.&,*= 8= 5->8.(&1= 5742.8(:.9>`= (43+1.(9= &3)= (-*&9.3,_= 8*<= ->9414,.89= +/. a.81_=

106 &7)*8`= #_`= &3)= ._= 5_= 7:38_= +33-_= 98= 7.2*78= <.9-= !3-&3(*)= 5*(.+.(.9>= +47= &8.).42>(*9*8= 8= 551.(&9.43=94=9-*=)*39.+.(&9.43=4+=#>(477-.?&*=&3)=$:898_=#41*(:1&7=!(414,>= ,a++-8++2_=

*'&:*7`= _= ,**/_= &793*79&:8(-= .2= ):301*3= F&1)= 89&'.1*= 84945*= ,*'*3= 3*:*= !.3'1.(0*= .3= )&8= !73C-7:3,8;*7-&19*3=;43=%7(-.)**3_=)_=-*_=*)_=&>*7.8(-*=0&)*2.*=)*7=F.88*38(-&+9*3_= C*71&,=57_=B7.*)7.(-=+*.1`=#33(-*3`= *72&3>_=

*'&:*7`= _`= &3)= #_= #*>*7_= ,**-_= 88+/= &3)= 18+-= 3&9:7&1= &':3)&3(*= 4+= &:949745-.(= &3)= 2>(4-*9*749745-.(= 47(-.)8= 574;.)*8= .38.,-9= .394= 3.974,*3= &3)= (&7'43= ,&.3= +742= +:3,&1= &884(.&9.43_=8*<=->9414,.89= +0* a,*38,,-_=

.71&3)&`=#_`=#_=_=*1488*`=5_=1&+&884`=B_=7.11.`=_=5*1+.3*`=$_=B&''.&3`=_= -.,343*`=_=.3*11.`=$_= *,7*94`= B_= '47*94`= _= 14??41.34`= &3)= _= *74994_= ,**0_= 3*++.(.*39= 5-4948>39-*8.8= .3= 9-*= #*).9*77&3*&3= 47(-.)= .24)47:2= &'479.;:2= .8= 2.7747*)= '>= 85*(.+.(= &884(.&9.43= 94= *(942>(477-.?&1=$:88:1&(*&*_=#41*(:1&7=!(414,>= +/ a.3+8/*._=

1*3`=#_`=_=1_=.422*7:5`=8_='_=4:,-*7`=&3)=_=_=%7.*3_=,**,_=7*=*'&(.3&(*&*=(42243=&3)= <.)*857*&)= *(942>(477-.?&1= &884(.&9*8= 4+= !:(&1>59:8= 85*(.*8= .3= :897&1.&3= +47*898= #>(477-.?&= +, a,.-8,.1_=

&)1*>`= _= +31*_= 84385*(.+.(.9>= 4+= >2'49.(= 3+*(9.43= .3= %7(-.)= #>(477-.?&_= 8*<= ->9414,.89= 03 a+*+/8]_=

&2&)&`=#_=+3-3_=9:).*3=3'*7=).*=#>0477-.?&=;43= .&1*41&=8*59*397.43&1.8 =$*.(-'_=+_=8=!.3=3*:*7=B&11= )*7= #>0477-.?&8'.1):3,= ):7(-= .397&7&).(&1*= $-.?42475-&_= <&5&3*8*= <4:73&1= 4+= 49&3>= +* a+/+8,++_=

&2&)&`= #_`= &3)= _= _= 8&0&2:7&_= +30-_= F:7?*18>2'.48*= ;43= .&1*41&= &19.88.2& = $*.(-'_= +_`= !.3*7= (-14745->11+7*.*3=%7(-.)**`=2.9=)*2= 41??*7897*3)*3=.1?= >2*34(-&*9*=(74(.(7*&8=*70_= 7_=(._=$*5_=.4-40:=03;._=*7_=.=.414,>= ,3 a,,18,-2_=

/479-`=!_`=A_= &)+.`=_=_= 4:1)`=%_=A&<&(-`=#_=_=422*7`=_=Z&:3*7`=&3)=0_8 _=#&.*7_=,**._=Z*74`= 43*`= 9<4`= 9-7**`= &3)= 5*7-&58= +4:7= 8= !3)48>2'.48.8= &3)= 9-*= ,&.3= &3)= 1488= 4+= 51&89.)8_= !3)4(>94'.48.8=1*11=$*8*&7(-= +/ a./38.02_=

47943`= ._= $_`= &3)= ._= 5_= 7:38_= +332_= #:19.51*8-489= +:3,.= &7*= 9-*= 2489= +7*6:*39= &3)= &':3)&39= *(942>(477-.?&1=9>5*8=.3=&=2.=*)= 89&3)=4+=54:,1&8=+.7= 8*:)498:,&=2*3?.*8.. = &3)= '.8-45= 5.3*=.3:8=2:7.(&9&_=8*<=->9414,.89= +-3 a--+8--3_=

47943`=._=$_`=&3)=._=5_=7:38_=,**+_=.-*=241*(:1&7=7*;41:9.43=.3=*(942>(477-.?&1=*(414,>a=5**0.3,= .394=9-*='1&(08'4=_=#41*(:1&7=!(414,>= +* a+2//8+21+_=

2-4+`= _= ,**-_= = )478.;*397&1= 2>(477-.?&1= 7449= .3= 9-*= &(-14745->114:8= #(.&5-.1&= 541>,>3&= .7.:7.)&(*&*_=#>(477-.?&= +- a-,18--,_=

7<.3`= #_= <_`= <_= <_= 4:,4:7*`= &3)= <_= 5_= F_= 5*&73&1*>_= ,**1_= 9*7489>1.8= 3:9&38 = %7(-.)&(*&*= -&8= &= 85*(.+.(=&884(.&9.43=<.9-=9<4= *7&94'&8.).:2 =74498&884(.&9*)=+:3,.=&(7488=.98=7&3,*=.3=*&89*73= :897&1.&_=#>(48(.*3(*= .2 a,-+8,-3_=

<4-3843`=5_`=<_=$_='*&0*`=&3)=5_=<_=$*&)_=,**,_=.7&38+*7= 4+= 7*(*39= 5-4948>39-&9*= .394= 2>(477-.?&1= 2>(*1.:2=4+=&3=:51&3)=,7&881&3)a=8-47989*72=7*85.7&947>=1488*8=&3)=&((:2:1&9.43=4+=18+._= 4.1=.414,>=]=.4(-*2.897>= -. a+/,+8+/,._=

107 <:14:`=._`=_=:7,-&7)9`= _= *'&:*7`=5_=*7;*.11*7`=1_=5&2*8.3`=&3)=#_=_=*1488*_=,**/_=#.=49745->= .3=47(-.)8a=.38.,-98=+742=&=(425&7&9.;*=89:)>=4+=,7**3=.3).;.):&18=&3)=3435-4948>39-*9.(= .3).;.):&18=4+= *5-&1&39-*7&=)&2&843.:2 _=8*<=->9414,.89= +00 a0-380/-_=

A&2.*380.`= B_= +32,_= '*8= 47,&3*8= ;S,S9&9.+8= ):= 4349745&= ->545.9->8 = '_= #S24.7*= )*= 1&= 4(.S9S= 8&9.43&1*=)*8=(.*3(*8=8&9:7*11*8=*9=#&9-S2&9.6:*8=)*=1-*7'4:7,= ,0 a+8.*_=

A*33*)>`= _= _`= _= 5_= ??4`= &3)= ._= 5_= 7:38_= ,**-_= .-*7*= .8= -.,-= 549*39.&1= +47= 9-*= +472&9.43= 4+= (42243= 2>(477-.?&1= 3*9<4708= '*9<**3= :3)*78947*>= &3)= (&345>= 97**8= .3= &= 2.=*)= *;*7,7**3=+47*89_=<4:73&1=4+=!(414,>= 3+ a+*1+8+*2*_=

A1*.3`=!_=+313_=5.*=&54(-742*3=B&7';&7.*9C9*3=)*8= 5.5&(9.8=&9747:'*38 = %BB#_=!!$_=%7(-.)**= -* a38+-_=

A1.2*84;&`= <_= ,**1_= $44988574:9.3,= .3= 2>(48-*9*749745-.(= 51&398a= 57*5&(0&,*)= 8>2'.48*8= 47= 4;*7(42.3,=2*7.89*2=1.2.9&9.43=8*<=->9414,.89= +1- a28+*_=

A7&:8*`=A_=,**2_=B742=(-147451&898=94=(7>59.(=51&89.)8a=*;41:9.43=4+=51&89.)=,*342*8=.3=5&7&8.9.(= 51&398_=1:77*39= *3*9.(8= /. a+++8+,+_=

A7*9?*7`=_=#_`=#_=_=.)&7943)4`='_=1_= 7:'.8-&`=<_=F_=5&9&+47&`=._=#_=?&74`=&3)=._=5_=7:38_=,***_= $*,.43&1=85*(.&1.?&9.43=4+= #&7(4)*8=8&3,:.3*& =!7.(&(*&*=43=&=8.3,1*=+:3,&1=8>2'.439=+742= 9-*= --.?454,43= *11*3&*= $-.?454,43&(*&*= 85*(.*8= (4251*=_= 2*7.(&3= <4:73&1= 4+= 49&3>= 21 a+1128+12,_=

A:78&7`=._=_`=&3)=_=5_=141*>_=,**-_=143;*7,*3(*=.3=)*+*38*=8>3)742*8=4+=>4:3,=1*&;*8=.3=9745.(&1= 7&.3+47*898_=.4(-*2.(&1=>89*2&9.(8=&3)=!(414,>= -+ a3,383.3_=

A:8&34`=_=+3++_= .&8974).&=*1&9& =&3)=.98=8>2'.49.(=&884(.&9.43=<.9-= 72.11&7.&=2*11*&_ =<4:73&1=4+=9-*= 1411*,*=4+=,7.(:19:7*`=03.;*78.9>=4+=.40>4= .a+80+_=

'&3(-.*7`=8_`=&3)=1_=8*:-&:8*7_=,**0_==85&9.&11>=*=51.(.9=24)*1=+47=(425*9.9.43=&243,=85*(.&1.898= &3)= ,*3*7&1.898= .3= &= -*9*74,*3*4:8= *3;.7432*39_= 33&18= 4+= 551.*)= 74'&'.1.9>= +0 a+-2/8 +.+*_=

'*&0*`= <_= $_= +33._= .-*= .414,>= 4+= #>(48 *9*749745-.(= &5745->9.(= 1&398_= 8*<= ->9414,.89= +,1 a+1+8,+0_=

'*&0*`= <_= $_= ,**._= #>(48-*9*749745-2*5.5&7&8.9.(= 51&39= .39*7&(9.438= <.9-= *(942>(477-.?&1= &3)= &7':8(:1&7=2>(477-.?&1=+:3,._=1:77*39=%5.3.43=.3=1&39=.414,>= 1a.,,8.,2_=

'*&0*`=<_=$_`=5_=<4-3843`=5_=_=5433*11>`= _=!_=#:(01*`='_=4))>`=&3)=5_=<_=$*&)_=,**._=8*9<4708=4+= 54<*7=&3)=.3+1:*3(*a=9-*=741*=4+=2>(477-.?&1=2>(*1.:2=.3=(4397411.3,=51&39=(422:3.9.*8= &3)= &,74*(48>89*2= +:3(9.43.3,_= 1&3&).&3= <4:73&1= 4+= 49&3>8$*;:*= 1&3&).*33*= 5*= 49&3.6:*= 2, a+*+08+*./_=

'45*?`=_`=B_=$4:88*9`=B_= _=-&<`=$_= _=-&<`=&3)=%_=$43(*_=,**2_=#.,7&9.43=14&)=.3=51&398a=741*=4+= 5411*3= &3)= 8**)= ).85*78&1= .3= -*9*74,*3*4:8= 1&3)8(&5*8_= <4:73&1= 4+= !;41:9.43&7>= .414,>= ,+ a,3.8-*3_=

#&.741)`=B_`=&3)=B_=F*'*7_=+3/*_=849.?=3'*7= *5-&1&39-*7& =&1'.348_=749451&82&= -3 a,1/8,11_=

108 #&7948`=B_`=#_=5:14723*`=._=&.11*7`=_=43+&39*`=_=B&((.4`=<_=B4:73*1`=#_=_=5:'4.8`=&3)=#_=_= *1488*_== ,**3_=1-&11*3,.3,=*(414,.(&1=7:1*8=+742=9*25*7&9*=7*,.438a=9<4=9745.(&1= &(-14745->114:8== 47(-.)8=7*(*.;*=(&7'43=+742=8&574'.(=+:3,._=.3=57*5_= = #&8:-&7&`= _`=&3)=A_=A&98:>&_=+323_=!++*(98=4+=#>(477-.?&1=B:3,.=43=**)8 *72.3&9.43=&3)=!&71>= 74<9-=4+=-=<&5&3*8*=.*77*897.&1=%7(-.)8_=(.*39.&= 479.(:19:7&*= -1 a--+8--1_=

#&8:-&7&`= _`= &3)= A_= A&98:>&_= +33._= 38.9:= &3)= .3= ;.974 = 85*(.+.(.9>= '*9<**3 = --.?4(943.&$ 855= &3)= #5.7&39-*8=8.3*38.8 =*78443=2*8=C&7=24*3&=#=.*'*789*.3= &7&=%7(-.)&(*&*_=8*<= ->9414,.89= +,1 a1++81+2_=

#&8:-&7&`= _`= A_= A&98:>&`= &3)= A_= &&2&,:(-._= +33-_= 49*39.&1= +47= >2'.48.8= 4+ = --.?4(943.&$841&3.= &3)=.3:(1*&9*= --.?4(943.&$ <.9-=**)8=4+= #5.7&39-*8=8.3*38.8 =;&7= &24*3& =.3=;.974 _=#>(414,.(&1= $*8*&7(-= 31 a1.081/,_=

#&99843`=F_=<_=+32*_= *7'.;47>=.3=$*1&9.43=94=1&39=8.974,*381439*39_=33:&1=$*;.*<=4+=!(414,>= &3)=>89*2&9.(8= ++ a++38+0+_=

#&:34:7>`= B_`= 5_= *7;*.11*7`= 1_= '*1&7,*`= <_= &_= 439&.11*7`= '_= C&3'489&1`= &3)= 1_= 5&2*8.3_= ,**1_= *&843&1`=)&.1>=&3)=).:73&1=;&7.&9.438=.3=9-*=89&'1*=(&7'43=.84945*=(42548.9.43=4+=(&7'43= ).4=.)*=7*85.7*)='>=97**=97:308=.3=&=)*(.):4:8=4&0=+47*89_=%*(414,.&= +/+ a,028,13_=

#(1472.(0`=#_=A_`=5_=B_=F-.,-&2`=&3)=<_=%8*.11_=,**._= #>(477-.?&1= ).;*78.9>= .3= 5-4948>39-*9.(= 9*77*897.&1=47(-.)8_=8*<=->9414,.89= +0- a.,/8.-2_=

#( :.7*`= A_= '_= ,**1_= 142243= *(942>(477-.?&1= 3*9<4708= 2&>= 2&.39&.3= 2434)42.3&3(*= .3= &= 9745.(&1=7&.3=+47*89_=!(414,>= 22 a/018/1._=

#(A*3)7.(0`=_='_`=<_=$_='*&0*`=&3)=5_=<_=$*&)_=,***_= >2'.49.(= ,*72.3&9.43= &3)= )*;*1452*39= 4+= 2>(48-*9*749745-.(=51&398=.3=3&9:7*a=97&38+*7=4+=(&7'43=+742=*(942>(477-.?&1=&1.==7*5*38= &3)= *9:1&= 5*3):1&= 94= 9-*= 47(-.)= 47&1147-.?&= 97.+.)& = 9-74:,-= 8-&7*)= ->5-&1= (433*(9.438_= 8*<=->9414,.89= +./ a/-38/.2_=

#*7(0=`= C_`= '_= F_= 1-&974:`= _= '*2&.7*`= #_= 8_= &.3,*`= _= :>82&38`= &3)= !_= B_= 2*98_= ,**2_= 5.;*78.+.(&9.43= 4+= 2>(48-*9*749745-.(= &3,.485*728a= !;.)*3(*= +742= :72&33.&(*&*_= 2(= !;41:9.43&7>=.414,>= 2a+8+0_=

#41.3&`=$_`= _=#&88.(499*`=&3)=<_=#_=.7&55*_=+33,_=5*(.+.(.9>=5-*342*3&=.3=2>(477-.?&1=8>2'.48.8a= (422:3.9>8*(414,.(&1= (438*6:*3(*8= &3)= 57&(9.(&1= .251.(&9.438_= &,*8= -/18.,-= .3 = #_= B_= 11*3`=*).947_=#>(477-.?&1=+:3(9.433.3,a=&3=.39*,7&9.;*=51&398+:3,&1=574(*88_=1-&52&3=&3)= &11`=8*<=&470`=0_=

#41;7&>`= #_`= _= <_= A47*8`= &3)= #_= F_= 1-&8*_= ,***_= 41>5->1>= 4+= 2>(4-*9*749745-.(= 47(-.)8= &3)= +:3(9.433&1= .3+1:*3(*8= 43= +147&1= &3)= 241*(:1&7= (-&7&(9*78_= &,*8= ..+8..1= .3 = A_= '_= F.1843= &3)=5_=_=#477.843`=*).9478_=#434(498a=8>89*2&9.(8=&3)=*;41:9.43_=1$%`=#*1'4:73*_=

#447*`=$_=._=+321_=.-*= *3*7&=4+=$-.?4(943.&8'.0*=B:3,.=8= 8(47-.?4(943.&`=*7&947-.?& `= 5:147-.?& `= 43.1.458.8 `=&3)= --.?4(943.& _=#>(49&=43= ,3 a3+833_=

#4>*784*3`= _`= _= *(0*7`= &3)= _= <_= 1*=&3)*7_= ,**+_= 7*= *(942>(477-.?&8= 247*= &':3)&39= 9-&3= &7':8(:1&7=2>(477-.?&8=.3=9745.(&1=-*&9-=+47*898=8*<=->9414,.89= +/* a/3+8/33_=

109 #>*78`=8_`=$_=_=#.99*72*.*7`=1_= _=#.99*72*.*7`= _=_=_=)&=B438*(&`=&3)=<_=A*39_=,***_=.4).;*78.9>= -4985498=+47=(438*7;&9.43=57.47.9.*8_=8&9:7*= .*- a2/-82/2_=

84;493>`=C_`=&_=&88*9`=_=!_=#.11*7`= _=5_=F*.'1*3`=_=7*2*7`='_=1.?*0`=&3)=_=574?)_=,**,'_='4<= -489=85*(.+.(.9>=4+=-*7'.;474:8=.38*(98=.3=&=9745.(&1=+47*89_=8&9:7*= .+0 a2.+82.._=

84;493>`=C_`=_=574?)`=_=!_=#.11*7`=#_=A:1+&3`=#_=<&3)&`=&_=&88*9`=&3)= _=5_=F*.'1*3_=,**0_=F->= &7*=9-*7*=84=2&3>=85*(.*8=4+=-*7'.;474:8=.38*(98=.3=9745.(&1=7&.3+47*898=(.*3(*= -+- a+++/8 +++2_=

%,:7&8.8:/.9&`=&_`=&3)=._=&:0&<&_=,**2&_= 5.5&(9.8=-*11*'47.3* =8-4<8=89743,=2>(477-.?&1=57*+*7*3(*= 94<&7)8= *(942>(477-.?&1= +:3,.= <.9-= (4397&89.3,= ,*4,7&5-.(= ).897.':9.438= .3= <&5&3_= #>(477-.?&= +2 a--+8--2_=

%,:7&8.8:/.9&`= &_`= &3)= ._= &:0&<&_= ,**2'_= .,-= 2>(477-.?&1= 85*(.+.(.9>= .3= &= <.)*857*&)= 2>(4-*9*749745-.(= 51&39`= :145-.&= ?411.3,*7. = %7(-.)&(*&*_= 2*7.(&3= <4:73&1= 4+= 49&3>= 3/ a3-831_=

%3,:*3*`=8_=_`=&3)=._=F_=A:>5*7_=,**,_=25479&3(*=4+=9-*=*(942>(477-.?&1=3*9<470=+47=8**)1.3,= 8:7;.;&1= &3)= *(942>(477-.?&= +472&9.43= .3= 7&.3= +47*898= 4+= 84:9-= 1&2*7443_= #>(477-.?&= +, a+-8+1_=

%9*74`= <_= ._`= <_= 5_= (0*72&3`= &3)= _= &>2&3_= ,**,_= 5.;*78.9>= &3)= -489= 85*(.+.(.9>= 4+= *3)45->9.(= --.?4(943.& 81.0*=+:3,.=+742=9745.(&1=47(-.)8_=2*7.(&3=<4:73&1=4+=49&3>= 23 a+2/,8+2/2_=

%9*74`=<_=._`=<_=5_=(0*72&3`=&3)=_=&>2&3_=,**._=5.++*7*3(*8=.3=2>(477-.?&1=57*+*7*3(*8='*9<**3= 9<4=9745.(&1=47(-.)8_=#41*(:1&7=!(414,>= +- a,-3-8,.*._=

%9*74`=<_=._`=_=&>2&3`= &3)=<_=5_=(0*72&3_=,**/_= C&7.&9.43= .3= 2>(477-.?&1= 5*7+472&3(*= .3= 9-*= *5.5->9.(=47(-.)= "41:23.&=;&7.*,&9& =.3=;.974a=9-*=549*39.&1=+47=3&9:7&1=8*1*(9.43_=!;41:9.43&7>= !(414,>= +3 a,38.-_=

%9*74`= <_= ._`= 8_= _= B1&3&,&3`= !_= _= *77*`= <_= 5_= (0*72&3`= &3)= _= &>2&3_= ,**1_= F.)*857*&)= 2>(477-.?&1= 85*(.+.(.9>= (477*1&9*8= 94= 2>(477-.?&1= +:3(9.43= .3= 9-*= 3*49745.(&1`= *5.5->9.(= 47(-.) =!43458.8=:97.(:1&7.4.)*8 =%7(-.)&(*&*_=2*7.(&3=<4:73&1=4+=49&3>= 3. a+3..8+3/*_=

*7*.7&`= <_= _= +33/_= &8= !=(-&3,*= &3)= 74<9-_= &,*8= +.18+2+= .3 = !_85_= (-:1?*= &3)= _= 1&1)<*11`= *).9478_=!(45->8.414,>=4+=-4948>39-*8.8_=57.3,*7`=8*<=&470_=

+*.+-4+*7`= _= F_= +323_= !;.)*3(*= +47= 1-14745->118= &3)= '&(0= 4+= ':9*.3= .3= $*499.&8$.):88;.8 = 1&89.)8_=.4(-*2.*=03)=->8.414,.*=5*7=+1&3?*3= +2. a//80+_=

-.11.58`=5_='_`=&3)=<_=F_= 7*,,_=,**+_=03(*79&.39>=.3=84:7(*=5&79.9.43.3,=:8.3,=89&'1*=.84945*8=;41= +1,`=5,=+1+`=,**+_=%*(414,.&= +,2 a-*.8-*._=

7.),*43`=_=#_`=_=<_=17.''`=#_=#_=1-&8*`=&3)=B_=8_=$&82:88*3_=,**/_= *3*7&=47(-.)&(*&7:2a=;41=.a= !5.)*3)74.)&*_=%=+47)=03.;*78.9>=7*88`=%=+47)`=0A_=

$&2.7*?`=_=$_`=_= 7&;*3)**1`=$_=_=.3,*7`=1_=$_=#&78-&11`=&3)=8_=!_=.*7(*_=,**1_=5&9.3,=9-*=47.,.3= 4+=9-*=%7(-.)&(*&*=+742=&=+488.1=47(-.)=<.9-=.98=5411.3&947_=8&9:7*= ..2 a+*.,8+*./_=

$&82:88*3`= _= 8_= +33/_= .*77*897.&1= %7(-.)8= := B742= **)= 94= #>(49745-.(= 1&39_= 1&2'7.),*= 03.;*78.9>=7*88_=

110 $*)+*73`= 5_= _= +312_= 3+*(9.43= '>= 72.11&7.&8#*11*&= &3)= 42*= B&(9478= ++*(9.3,= 4898$*8.89&3(*= &3)=*;*7.9>=4+=5.8*&8*_=B47*897>= /+ a+,+8]_=

$*>*887.*94`=_`=_=#4:89&+&`=&3)=5_=-&99&(-&7>&_=,**2_=#:19.51*=,*3*8=4+=&55&7*39=&1,&1=47.,.3= 8:,,*89=(.1.&9*8=2&>=43(*=-&;*='**3=5-4948>39-*9.(_=1:77*39=.414,>= +2 a3/0830,_=

$.(-&7)`= B_`= _= #.1149`= #_= &7)*8`= &3)= #_= _= *1488*_= ,**/_= 5.;*78.9>= &3)= 85*(.+.(.9>= 4+= *(942>(477-.?&1= +:3,.= 7*97.*;*)= +742= &3= 41)8,74<9-= #*).9*77&3*&3= +47*89= )42.3&9*)= '>= 9:*7(:8=.1*= _=8*<=->9414,.89= +00 a+*++8+*,-_=

$.(-&7)`=B_`=_=_=#47*&:`=#_=_=*1488*`=&3)=#_= &7)*8_=,**._=5.;*78.9>=&3)=+7:.9.3,=5&99*738=4+= *(942>(477-.?&1=&3)=8&574'.(=+:3,.=.3=&3=41)8,74<9-=#*).9*77&3*&3=+47*89=)42.3&9*)='>= 9:*7(:8=.1*= ='_=1&3&).&3=<4:73&1=4+=49&3>8$*;:*=1&3&).*33*=5*=49&3.6:*= 2, a+1++8+1,3_=

$.(-&7)`=B_`=#_=_=*1488*`=&3)=#_= &7)*8_=,**3_=!89&'1.8-2*39=4+= 9:*7(:8=.1*= =8**)1.3,8=:3)*7=9<4= *&71>= 8:((*88.43&1= *7.(&(*4:8= 8-7:'8= .3= &= #*).9*77&3*&3= *(48>89*2a= 741*= 4+= 2>(477-.?&1= +:3,._=B!#8=#.(74'.414,>=]=!(414,>= 02 a+.8,._=

$.;.*7*`=._`=_= _=5.*)-.4:`=#_=5.&'&9*`= _=*39-.1&7&8:`=A_=8&9&7&/&3`=_=C*7'*0*3`=_=:>(0`=_= 57*>+:8`= _=*3&`=&3)=_=#_=&_=,**1_= *3*9.(=).;*78.9>=4+=*(942>(477-.?&1=&8.).42>(*9*8= +742=+7.(&3=&3)=3).&3=9745.(&1=7&.3=+47*898_=#>(477-.?&= +1 a.+/8.,2_=

$4'*79`=1_`=#_=%_=&3(&1`=_=8*>`=&3)=1_='&334:_=,**/_=F-*&9=1*&+=5-4948>39-*8.8=1488=):*=94=1*&+= 7:89`=<.9-=7*85*(9=94=1*8.43=)*;*1452*39=&3)=1*&+=3.974,*3=89&9:8_=8*<=->9414,.89= +0/ a,,18 ,.+_=

$4'.3`=#_=+323_=!;41:9.43=):3*=545:1&9.43=) 5.54,.:2=&5->11:2 a=:3*=S3.,2*=)S;4.1S*=1&-.*78=)*= 1&=4(.S9S=B7&3O&.8*=)%7(-.)45-.1.*= 0a,/.8,0*_=

$43(*`= %_`= &3)= #_= A.705&97.(0_= ,**+_= F-*3= 84:7(*8= '*(42*= 8.308a= #.,7&9.43&1= 2*19)4<3= .3= -*9*74,*3*4:8=-&'.9&98_=!;41:9.43= // a+/,*8+/-+_=

&)4;80>`=%_=+30/_=%7(-.)**3=.2=*.,*3*3= &79*3_='C`=#33(-*3`= *72&3>_= = &1)&77.&,&`=<_=B_`=B_=<_=$_=.&>147`=_=<_=A**1.3,`=&3)=._=1&;&1.*782.9-_=,**+_=5.34+1&,*11&9*=3:(1*&7= 0= 7$8= 5->14,*3>= 8:,,*898= 2:19.51*= 51&89.)= 1488*8= &3)= 7*51&(*2*398_= <4:73&1= 4+= #41*(:1&7=!;41:9.43= /- a,*.8,+-_=

&12.&`=_=+320_=1-14745->118+7**=+472=4+= 5.5&(9.8=-*11*'47.3*= %7(-.)&(*&*=.3=!=B.31&3)_=33&18= 4+=49&3>=B*33.(.= ,- a.38/1_=

&12.&`= _= +323&_= B*&9:7*8= 4+= *3)42>(477-.?&1= .3+*(9.43= 4+= (-14745->118+7**= &3)= ,7**3= +4728= 4+= 5.5&9.8=-*11*'47.3*= %7(-.)&(*&*_=33&18=4+=49&3>=B*33.(.= ,0 a+/8,0_=

&12.&`= _= +323'_= *3*7&1= 2475-4714,>= &3)= &3&942>= 4+= (-14745->118+7**= &3)= ,7**3= +4728= 4+= 5.5&(9.8=-*11*'47.3* =%7(-.)&(*&*_=33&18=4+=49&3>=B*33.(.= ,0 a3/8+*/_=

(&((-.`=$_`= _=5*&3,*1.8`=&3)=_='&3?&7&_=+33*_=114?>2*=C&7.&9.43=&243,=&3)=<.9-.3=++=%7(-.)8= 5*(.*8= B&2= %7(-.)&(*&*`= <.9-= 5*(.&1= $*+*7*3(*= 94= >'7.).?.3,= 59.9:)*_= *3*9.(&= 2+ a+.-8+/*_=

111 (&55&9.((.`=1_`=&3)= _=(&55&9.((._=+332_= 5.5&(9.8=2.(475->11& =! $ $5.=F$.Z=1:8:8=748*&= %%_= .3 = 1&-.*78= )*= 1&= 4(.9S9S= B7&3O&.8*= )%7(-.)45-.1.*`= +*7*8= /4:73S*8= $*3(4397*8= %7(-.)45-.1*8=$-3*8815*8_='>43_=

(-<*3)*7`=<_`=B_= 4++2&3`=<_=_=%-174,,*`=&3)=&_=-&(-&78 .11_=,**._=$:'.8(4=<.9-4:9=9-*=1&1;.3= (>(1*=.2574;*8=9-*=(&7'43=*++.(.*3(>=4+=)*;*145.3,=,7**3=8**)8_=8&9:7*= .-, a113812,_=

*.)*3+&)*3`= _=+312_=%7(-.)= *3*7&=.3=.-&.1&3)=Ca=8*499.4.)*&*='.3)1_=5&380=49&3.80=B47*3.3,_=

*1488*`=#_=_=,***_='&=8>2'.48*a=897:(9:7*8=*9=+43(9.438`=71*=S(414,.6:*=*9=S;41:9.+_=C:.'*79`=&7.8_=

*1488*`=#_=_`=!_=&:)4.3`=&3)=_=C&3)*304473-:>8*_=,**.&_=>2'.49.(=2.(7447,&3.828`=&=0*>=+47= *(414,.(&1=8:((*88=&3)=5749*(9.43=4+=51&398_=14259*8=$*3):8=.414,.*8= -,1 a0-380.2_=

*1488*`= #_= _`= $_= &:*7`= &3)= _= #4>*784*3_= ,**,&_= &8&1= ->2*342>(*9*8= '*143,.3,= 94= 9-*= *'&(.3&(*&*=&7*=*(942>(477-.?&1=43=9*25*7&9*=)*(.):4:8=97**8_=8*<=->9414,.89= +// a+2-8 +3/_=

*1488*`= #_= _`= _= B&((.4`= _= (&55&9.((.`= &3)= _= 43+&39*_= ,**.'_= 1-14745->114:8= &3)= &(-14745->114:8=85*(.2*38=4+= 5.5&(9.8=2.(745->11& =8*499.*&*`=%7(-.)&(*&*=&7*=&884(.&9*)= <.9-=*(942>(477-.?&1=8*5942>(*9*8`=.3(1:).3,=97:++1*8_=#.(74'.&1=!(414,>= .1 a.+08.,0_=

*1488*`=#_=_`=B_=$.(-&7)`=6_= _= *`=&3)=_=F_=.2&7)_=,**0_=#>(477-.?&1=3*9<4708a=)*8=1.&.8438= )&3,*7*:8*8=.7*3)8=.3=!(414,>=]=!;41:9.43= ,+ a0,+80,2_=

*1488*`=#_=_`=_=*9&74`=B_= 1&9&7)`=B_= $.(-&7)`= 1_=07(*1&>`=&3)=#_=F*.88_=,**1_=*'&(.3&1*8=&7*= (42243=2>(477-.?&1=&884(.&9*8=4+=!7.(&(*&*_=8*<=->9414,.89= +1. a20.8212_=

*1488*`= #_= _`= #_= F*.88`= <_= '_= <&3>`= &3)= _= ..11.*7_= ,**,'_= 1422:3.9.*8= &3)= 545:1&9.438= 4+= 8*'&(.34.)= '&8.).42>(*9*8= &884(.&9*)= <.9-= 9-*= &(-14745->114:8= 47(-.)= $*499.&= 3.):88&;.8 = '_='1#=$.(-_=&3)=3*.,-'4:7.3,=97**=*(942>(477-.?&*_=#41*(:1&7=!(414,>= ++ a+2-+8+2.._=

-*++*7843`=$_=_=,**0_=:7;.;&1=(4898=4+=&):19=)472&3(>=&3)=9-*=(43+4:3).3,=.3+1:*3(*=4+=8.?*=.3= 1&)>8=81.55*7=47(-.)8`=,*3:8 =>57.5*).:2 _=%.048= ++/ a,/-8,0,_=

-*++*7843`=$_=_`= ._=A:11`=&3)=A_=.&1._=,**/&_=):19= <-41*851&39= )472&3(>= .3):(*)= '>= 897*88= .3= 143,81.;*)=47(-.)8_=!(414,>= 20 a-*338-+*._=

-*++*7843`=$_=_`=._=A:11`=&3)=A_=.&1._=,**0_=5*24,7&5-.(=7*85438*=94=8-&).3,=&3)=)*+41.&9.43=.3= 9<4=<44)1&3)=47(-.)8_=B41.&= *4'49&3.(&= .+ a3/8+*0_=

-*++*7843`=$_=_`=5_='_=.&>147`=#_=F*.88`=_= &73.(&`=#_=A_=#(1472.(0`=_=)&28`= _=#_= 7&>`=<_=F_= #(B&71&3)`=._=A:11`=A_=.&1.`=._=&:0&<&`=._=A&<&-&7&`=A_=#.>48-.`=&3)=&_=_='**_=,**1_=.-*= *;41:9.43&7>= -.8947>= 4+= 2>(477-.?&1= 85*(.+.(.9>= &243,= 1&)>8= 81.55*7= 47(-.)8_= !;41:9.43= 0+ a+-2*8+-3*_=

-*++*7843`= $_= _`= #_= F*.88`= ._= A:11`= &3)= 5_= '_= .&>1478_= ,**/'_= .,-= 85*(.+.(.9>= ,*3*7&11>= (-&7&(9*7.?*8= 2>(477-.?&1= &884(.&9.43= .3= 7&7*= 1&)>8= 81.55*7= 47(-.)8`= ,*3:8= >57.5*).:2 _= #41*(:1&7=!(414,>= +. a0+-80,0_=

.2&7)`=_=F_`=&3)=5_=#_=5:7&11_=,**._=#>(477-.?&1=3*9<4708a=&=7*;.*<=4+=9-*.7=*=9*39`=+:3(9.43`=&3)= .25479&3(*_=1&3&).&3=<4:73&1=4+=49&3>8$*;:*=1&3&).*33*=5*=49&3.6:*= 2, a++.*8++0/_=

112 .2&7)`=_=F_`=5_=_=*77>`=#_=5_=<43*8`=5_=5_=#>741)`= 5_= #_= 5:7&11`= &3)= $_= #41.3&_= +331_= 8*9= 97&38+*7=4+=(&7'43='*9<**3=*(942>(477-.?&1=97**=85*(.*8=.3=9-*=+.*1)_=8&9:7*= -22 a/138/2,_=

2*)1*>`=#_=_`=._=!_=5&<843`=<_=_=142894(0`='_=_=5434;&3`=5_=!_=-*77.11`=1_=_=1440`=&3)=<_=$_= !-1*7.3,*7_= +33+_= *&843&1= 1&7'43= 84945*= 5.8(7.2.3&9.43= .3= &= 7&881&3)= 1422:3.9>_= %*(414,.&= 2/ a-+.8-,*_=

2.9-`=_=!_`=&3)=5_=<_=$*&)_=,**2_=#>(477-.?&1=>2'.48.8`=-=*).9.43_=(&)*2.(=7*88`='43)43`=0A_=

27*(.:`=!_=_`=&3)=$_=_=1:77&-_=+323_=>2'.49.(=,*72.3&9.43=4+=8**)=4+=9*77*897.&1=47(-.)8=4+=8479-= 2*7.(&=&3)=!:745*_='.3)1*>&3&= +a08+/_=

4>7.30.`= 8_= +321_= %3= 9-*= 5*7.4).(.9>= .3= 9-*= +14<*7.3,= 4+= 5.54,.:2= &5->11:2 = %7(-.)&(*&*_= #*247&3)&=4(.*9&8=74=B&:3&=*9=B147&=B*33.(&= 0- a0-81,_=

9*5-*38`=_=_`=F_=<_=:9-*71&3)`=&3)=$_=_=B7*(01*943_=+333_=F-&9=.8=9-*=11**=*++*(9=%.048= 21 a+2/8 +3*_=

94:,&&7)`= <_= ,***_= $*,:1&9478= &3)= 7*,:1&9.43= 4+= 1*,:2*= 7449= 34):1*= )*;*1452*39_= 1&39= 5->8.414,>= +,. a/-+8/.*_=

:22*7-&>*8`=C_=_=+3/+_=F.1)=47(-.)8=4+=7.9&.3=<.9-=&=0*>=94=9-*=85*(.*8_=1411.38`='43)43_=

.&>147`= 5_= '_`= &3)= ._= 5_= 7:38_= +331_= 3)*5*3)*39`= 85*(.&1.?*)= .3;&8.438= 4+= *(942>(477-.?&1= 2:9:&1.82= '>= 9<4= 3435-4948>39-*9.(= 47(-.)8_= 74(**).3,8= 4+= 9-*= 8&9.43&1= (&)*2>= 4+= (.*3(*8=4+=9-*=03.9*)=9&9*8=4+=2*7.(&= 3. a./+*8./+/_=

.&>147`= 5_= '_`= &3)= ._= 5_= 7:38_= +333_= 45:1&9.43`= -&'.9&9= &3)= ,*3*9.(= (477*1&9*8= 4+= 2>(477-.?&1= 85*(.&1.?&9.43=.3=9-*=(-*&9.3,=47(-.)8= 47&1147-.?&=2&(:1&9&= &3)= _=2*79*38.&3& _= #41*(:1&7= !(414,>= 2a+1+38+1-,_=

.&>147`=5_='_`=._=5_=7:38`=&3)=_=_= 4),*8_=,**._=!;.)*3(*=+47=2>(477-.?&1=7&(*8=.3=&=(-*&9.3,= 47(-.)_=74(**).3,8=4+=9-*=$4>&1=4(.*9>=4+='43)43=*7.*8=8.414,.(&1=(.*3(*8= ,1+ a-/8.-_=

.&>147`=5_='_`=._=5_=7:38`=<_=$_='*&0*`=&3)=5_=<_=$*&)_=,**,_=#>(477-.?&1=85*(.+.(.9>=&3)=+:3(9.43=.3= 2>(48-*9*749745-.(= 51&398_= &,*8= -1/:.+-= .3 = ;_= )_= _= #_= &3)*78= `= *).947_= #>(477-.?&1= !(414,>_=57.3,*7`=*71.3`= *72&3>_=

.&>147`= 5_= '_`= ._= 5_= 7:38`= ._= #_= ?&74`= &3)= _= _= 4),*8_= ,**-_= 5.;*7,*3(*= .3= 2>(477-.?&1= 85*(.&1.?&9.43= <.9-.3= **=&1*(97.8= 85.(&9& = %7(-.)&(*&*`= &= 3435-4948>39-*9.(= )*8*79= 47(-.)_= 2*7.(&3=<4:73&1=4+=49&3>= 3* a++028++13_=

.*)*7844`= '_`= _= *11*9`= 0_= A41/&1,`= &3)= #_= _= *1488*_= ,**1&_= &7&11*1= *;41:9.43&7>= 5&9-8= 94= 2>(4-*9*749745->= .3= :3)*78947*>= !7.(&(*&*= &3)= %7(-.)&(*&*a= *(414,.(&1= *;.)*3(*= +47= 2.=49745->=.3=>741*&*_=%*(414,.&= +/+ a,*08,+1_=

.*)*7844`='_`=._=:;.`=A_=*&;*7`=&3)=0_=A41/&1,_=,**1'_= !(942>(477-.?&1= +:3,.= 4+= 9-*= *>(-*11*8a= ).;*78.9>= 5&99*738= &3)= -489= 8-.+98= +742= 9-*= 3&9.;*= :&9*7.458.8= 8*>(-*11&7:2 = 5.59*74(&75&(*&*=&3)= !398.&='./:,& =1&*8&15.3.&(*&*=94=9-*=.3974):(*)= :(&1>59:8=74':89&= #>79&(*&*`=':9=349= .3:8=(&7.'*& =.3&(*&*_=8*<=->9414,.89= +1/ a-,+8---_=

113 .*)*7844`= '_`= ._= :;.`= A_= *&;*7`= &3)= _= &&7_= ,**1(_= !(942>(477-.?&8= 4+ = 4197.(.& = &3)= 4197.(.*11& = >2*34(-&*9&1*8`='&8.).42>(49&=43=(&*8&15.3.&(*&*`=).59*74(&75&(*&*=&3)=2>79&(*&*=.3= 8*>(-*11*8_=#>(414,.(&1=74,7*88= 0a+*+8+*1_=

.*7&8-.9&`= ._= +32/_= B:3,.= .3-&'.9.3,= <.1)= 47(-.)8= .3= <&5&3= _= = 8>2'.49.(= *=5*7.*2*39= <.9-= 72.11&7.&=2*11*& =&3)= .&1*41&=8*59*397.43&1.8_ =.7&38&(9.438=4+=9-*=#>(414,.(&1=4(.*9>=4+=<&5&3= ,0 a.18/-_=

.*7&8-.9&`=._`=&3)=_=1-:2&3_=+321_=B:3,.=.3-&'.9.3,=<.1)=47(-.)8=.3=<&5&3=C_= 72.11&7.&=9&'*8(*38 `= &=3*<=8>2'.439=4+= .&1*41&=8*59*397.43&1.8 _=.7&38&(9.438=4+=9-*=#>(414,.(&1=4(.*9>=4+=<&5&3= ,2 a+./8+/._=

.7:)*11`= _= _`= _= ._= $>,.*<.(?`= &3)= $_= '_= !)243)8_= ,**-_= 8.974,*3= &3)= (&7'43= 89&'1*= .84945*= &':3)&3(*8= 8:55479= 9-*= 2>(48-*9*749745-.(= 3&9:7*= &3)= -489885*(.+.(.9>= 4+= (*79&.3= &(-14745->114:8=51&398_=8*<=->9414,.89= +0* a-3+8.*+_=

.:'&`= Z_`= _= A_= '.(-9*39-&1*7`= Z_= 18.39&1&3`= Z_= 8&,>`= &3)= A_= ?*39*_= +330_= '488= 4+= (-14745->118`= (*88&9.43=4+=5-4948>39-*9.(=1%,=&88.2.1&9.43=&3)=7*85.7&9.43=.3=9-*=54.0.14(-14745->114:8= 51&39= 0*745->9&=8(&'7.)& =):7.3,=)*8.((&9.43_=->8.414,.&=1&39&7:2= 30 a-2-8-22_=

.:73':11`=#_= _`=5_=541*>`=&3)=5_=<_=&&9*8_=+33-_=.-*=5>3&2.(8=4+=-4948>39-*9.(=((1.2&9.43=94= 1-&3,*8=.3='.,-9=I:&39.9>=&3)=I:&1.9>=.3=-=:897&1.&3=$&.38B47*89=.7**=5*(.*8_=%*(414,.&= 3. a,+28,,2_=

.<.*,`=_=5_`=5_=#_=5:7&11`=&3)=_=F_=.2&7)_=,**1_=!(942>(477-.?&1= +:3,&1= 8:((*88.43= .3= 2.=*)= 9*25*7&9*=+47*898_=8*<=->9414,.89= +10 a.-18..1_=

02&9&`= _= +331_= 3= ;.974= ,*72.3&9.43= 4+= 7>9-747(-.8= 4(-4'.*38.8 = %7(-.)&(*&*= .3= 9-*= 57*8*3(*= 4+= >45->11:2=8-.2*1. `=&3=*(942>(477-.?&1=+:3,:8_=#>(48(.*3(*= -2 a-//8-/1_=

02&9&`= _= +332&_= !3= ;.974 = >2'.49.(= 884(.&9.43= 4+= &3= (-14745->114:8= %7(-.)`= 7>9-747(-.8= 4(-4'.*38.8 `= <.9-= %7(-.)= &3)= 8438%7(-.)= B:3,._= #*24.78= 4+= 9-*= +&(:19>= 4+= ,7.(:19:7*`= A&,48-.2&=03.;*78.9>= -. a318+*1_=

02&9&`= _= +332'_= = 3*<= '.414,.(&1= +:3(9.43= 4+= -..9&0*= 2:8-7442`= *39.3:1&= *14)*8 `= .3= &= 2>(4-*9*749745-.(=47(-.) `=7>9-747(-.8=4(-4'.*38.8 _=#>(48(.*3(*= -3 a2/822_=

03,*7`=B_=+2.*_=*.97C,*=?:7=A*3393.8=)*7=5&7&8.9.8(-*3=+1&3*3_=33&1*3=)*8=F.*3*7=#:8*:28=)*8= 8&9:7,*8(-.(-9*= ,a+-80*_=

C&3=)*7=1.3,*1`=8_=_=+33/_=3=&91&8=4+=47(-.)=5411.3&9.438!:745*&3=47(-.)8_=&10*2&`=$499*7)&2_=

C&3= ..*3)*7*3`= _= _= +331_= *3*7&1.898`= 85*(.&1.898= &3)= 9-*= *;41:9.43= 4+= 5-*349>5.(= 51&89.(.9>= .3= 8>25&97.(=545:1&9.438=4+=).89.3(9=85*(.*8_=!;41:9.43= /+ a+-1,8++-*_=

C&3=C&1*3`='_=+311_=.-*=$*)=I:**3_=.-*=2*7.(&3=8&9:7&1.89= +++ a2*382+*_=

C.411*`=1_`=#_='_=8&;&8`=5_=C.1*`=!_=A&?&04:`=1_=B479:3*1`=_= :22*1`=&3)=!_= &73.*7_=,**1_='*9=9-*= (43(*59=4+=97&.9='*=+:3(9.43&1=%.048= ++0 a22,823,_=

F&7(:5`= <_= _= +32/_= $-.?&39-*11&8 &7)3*7.= %7(-.)&(*&*`= 98= --.?4(943.&$ !3)45->9*= &3)= 1148*= 884(.&9.43= <.9-= *1&1*:(&= :3(.3&9& = #>79&(*&*= .3= F*89*738:897&1.&_= 8*<= ->9414,.89= 33 a,1-8,2*_=

114 F&7(:5`=<_= _`=&3)=_= _=_=.&1'49_=+301_=*7+*(9=9&9*8=4+=$-.?4(943.&8$=884(.&9*)=<.9-=%7(-.)8_= 8*<=->9414,.89= 00 a0-+8]_=

F&9*72&3`=$_=<_`=&3)=#_=_=.)&7943)4_=,**2_=5*(*59.43=&'4;*`=)*(*59.43='*14(477-.?&1='.414,>=4+=47(-.)8_=<4:73&1=4+=!=5*7.2*39&1=49&3>= /3 a+*2/8+*30_=

F*''`=1_=%_`=5_=5_=(0*71>`=#_=_=#(**0`=&3)=#_=<_=5434,-:*_=,**,_=->14,*3.*8=&3)=(422:3.9>= *(414,>_=33:&1=$*;.*<=4+=!(414,>=&3)=>89*2&9.(8= -- a.1/8/*/_=

F*.88`=#_`=#_=_=*1488*`=A_= _=$*=*7`=_=07'&3`=&3)=B_=%'*7<.301*7_=,**._=*'&(.3&1*8a=&=-.9-*794= 4;*71440*)=(482=4+=-*9*74'&8.).42>(*9*8=<.9-=&='74&)=2>(477-.?&1=549*39.&1_=#>(414,.(&1= $*8*&7(-= +*2 a+**-8+*+*_=

F-.97.),*`= _`= &3)= 5_= 4:9-<479-_= ,**/_= #>(477-.?&1= 8>2'.4398= 4+= 9-*= 9*77*897.&1= 47(-.)= >57.5*).:2=+&8(.(:1&9:2 _=*1'>&3&= ,0 a-,28--._=

F.11.&28`=<_= _= _`=&3)=<_=B_=B&77&7_=+33*_=1439741=4+=&71*>=$449=$*85.7&9.43_=->8.414,.&=1&39&7:2= 13 a,/38,00_=

F.11.&2843`=5_= _`=#_=<_= &7)3*7`=_=7*.8*7`=5_=<_=#447*`=A_=$&3,&(-&7.`=&3)=$_=<_=#_=F.1843_=+33._= .-*= !;41:9.43&7>= %7.,.3= 4+= 9-*= -/= A'= 1.7(:1&7= 58= 4+= 1&824).:28B&1(.5&7:2= 8= 8*<= !;.)*3(*= :554798= &= 488.'1*= $-4)45->9*= 3(*897>_= #41*(:1&7= ]= *3*7&1= *3*9.(8= ,.- a,.38,/,_=

F.1843`=$_=<_=#_`=_=F_=5*33>`=_=$_=7*.8*7`=A_=$&3,&(-&7.`=A_=$4'*798`=_=$4>`=_=F->9*`=#_=97&9-`= 5_=<_=#447*`=_=F_=#447*`=&3)=5_= _=F.11.&2843_=+330_=14251*9*=,*3*=2&5=4+=9-*=51&89.)8 1.0*= 58= 4+= 9-*= 2&1&7.&= 5&7&8.9*= 1&824).:2= +&1(.5&7:2_= <4:73&1= 4+= #41*(:1&7= .414,>= ,0+ a+//8+1,_=

F7.,-9`=_=+30/_=.-*=39*757*9&9.43=4+=45:1&9.43897:(9:7*= '>=B89&9.89.(8=<.9-= 5*(.&1=$*,&7)=94= >89*28=4+=#&9.3,_=!;41:9.43= +3 a-3/8.,*_=

&&,&2*`= ._`= ._= B:10.-&7:`= #_= &&2&94`= _= :?:0.`= &3)= A_= <&8*_= ,**2&_= )*39.+.(&9.43= 4+= &= 2>(477-.?&1=+:3,:8=.3= 5.54,.:2=748*:2 =%7(-.)&(*&*=+742=2475-414,.(&1=(-&7&(9*7.89.(8= 4+='&8.).42&9&_=#>(48(.*3(*= .3 a+.18+/+_=

&&,&2*`= ._`= #_= &&2&94`= _= :?:0.`= &3)= A_= <&8*_= ,**2'_= 1*7&94'&8.).&(*&*= 2>(477-.?&1= +:3,.= .841&9*)=+742=3435-4948>39-*9.(=47(-.)= -&2&*,&8974).&=8.040.&3& _=#>(477-.?&= +2 a318+*+_=

&&2&94`= #_`= &3)= A_= <&8*_= ,**2_= 3974):(9.43= 4+= &8>2'.49.(&11>= 5745&,&9*)= 8**)1.3,8= 4+= 1*5-&1&39-*7&= +&1(&9&= %7(-.)&(*&*= .394= 3&9:7&1= -&'.9&9= &3)= .3;*89.,&9.43= 4+= (4143.?*)= 2>(477-.?&1=+:3,._=!(414,.(&1=$*8*&7(-= ,- a-,38--1_=

=&&2&94`=#_`=._=&&,&2*`=_=:?:0.`=&3)=A_=<&8*_=,**/_=841&9.43=&3)=.)*39.+.(&9.43=4+=2>(477-.?&1= +:3,.= &884(.&9.3,= <.9-= &3= &(-14745->114:8= 51&39`= 5.54,.:2= 748*:2= %7(-.)&(*&*_= #>(48(.*3(*= .0 a1-811_=

&:<&824735.9&0`= ._`= ._= C.(-.98449-430:1`= #_= .&39.(-&74*3`= _= 1-**;&)-&3&7&0`= &3)= _= $&9(-&)&<43,_=,**0_=5.;*78.9>=4+=!(942>(477-.?&1=+:3,.=43=5.59*74(&75&(*&*=.3=.-&.1&3)_= <4:73&1=4+=.414,.(&1=(.*3(*8= 0a+*/38+*0._=

115 Z*12*7`=1_=5_`=&3)=$_=_=1:77&-_=+33/_=!;.)*3(*=+47= &= B:3,&1= '.&.843= '*9<**3= 47&1147-.?&= 97.+.)& = %7(-.)&(*&*= &3)= .3:881439479&= .3&(*&*_= 1&3&).&3= <4:73&1= 4+= 49&3>8$*;:*= 1&3&).*33*=5*=49&3.6:*= 1- a20,8200_=

Z*991*7`='_=F_`=<_=-&72&`=&3)=B_=8_=$&82:88*3_=,**-_=#>(477-.?&1=).;*78.9>_=&,*8=,*/8,,0= .3 =A_=F_= 5.=43`=_=_=A*11`=$_='_=&77*99`=&3)=_=<_=17.''`=*).9478_=%7(-.)=(438*7;&9.43_=8&9:7&1= .8947>= :'1.(&9.438`=A49&=A.3&'&1:`=&'&-`=#&1&>8.&_=

Z.22*7`= A_`= 8_= _= >3843`= _= *'&:*7`= !_= _= 11*3`= #_= B_= 11*3`= &3)= 5_= <_= $*&)_= ,**1_= F.)*= ,*4,7&5-.(&1= &3)= *(414,.(&1= ).897.':9.43= 4+= 3.974,*3= &3)= (&7'43= ,&.38= +742= +:3,.= .3= 5>7414.)8=&3)=243497454.)8=!7.(&(*&*=&3)=.3=47(-.)8_=8*<=->9414,.89= +1/ a+008+1/_=

Z.22*7`= A_`= 1_= #*>*7`= &3)= _= *'&:*7_= ,**2_= .-*= *(942>(477-.?&1= 85*(.&1.89= 47(-.)= 47&1147-.?&= 97.+.)& =.8=&=5&79.&1=2>(48-*9*749745-_=8*<=->9414,.89= +12 a-3/8.**_

116 79.(1*= 79.(1*= =

= _= `=_8_= `=_= `=_= `=_= =]=_8_= _= ;.)*3(*= +742= 545:1&9.43= ,*3*9.(8= 9-&9= 9-*= *(942>(477-.?&1= '&8.).42>(*9*= &((&7.&= &2*9->89.3& =.8=&3=&(9:&1=2:19.8-489=8>2'.439_ =41*(:1&7= (414,> `=,**2`=;41=+1=.88:*=+,`=55= ,2,/8,2-2_= = &= 85S(.+.(.9S= )*= 1&= 8>2'.48*= *(942>(47-.?.*33*= *89= 97T8= ;&7.&'1*`= (*79&.3*8= *85T(*8= 5*:;*39= U97*= 85S(.+.6:*8`= (422*= :88:1&= 1&7.(.3& = &884(.S*= &:= S1T?*= *9= )&:97*8= 8439= (438.)S7S*8=(422*=,S3S7&1.89*8=(&7=43=1*8=974:;*=A=574=.2.9S=)*=).;*78*8=*88*3(*8=)&7'7*8_= *=89&9:9=)*=,S3S7&1.89*=*89=)43(=,S3S7&1*2*39=)S+.3.=8:7=1&='&8*=)*=(7.9T7*8=2475-414,.6:*8_= 7=(*79&.3*8=S9:)*8=7S(*39*8=24397*39=1*8=1.2.9*8=)*=(*99*=7*(433&.88&3(*=*3=7S;S1&39=)*8=(&8= )*= 85S(.&9.43= (7>59.6:*`= (422*= (-*?= *34(4((:2= ,*45-.1:2 _= &38= 1&= 2*8:7*= 41= 1&= 343= 85S(.+.(.9S= )*= (*99*= 8>2'.48*= 5*72*9= )*=51.6:*7= 1*8= 97&38+*798= )*= (&7'43*= *397*= *85T(*8= )&7'7*8= *9= 1&= +472&9.43= )*= 7S8*&:= (422:3= 2>(47-.?.*3`= *9= 6:*= 1*= +43(9.433*2*39= )*= (*8= 5488.'1*8= 7S8*&:== +&.9= )S'&9`= 1&= )S243897&9.43= )*= 1&= 343885S(.+.(.9S= )-9*= )*;*3&.9= 3S(*88&.7*_= 4:8=&;438=)43(=(-4.8.=)*=7S543)7*=A=(*99*=6:*89.43`=A=1&.)*=)*8=2U2*8=4:9.18=6:.= 7S;T1*39= 1*8= (&8= )*= 85S(.&9.43= (7>59.6:*=a= 1&= ,S3S9.6:*= *9= 1S9:)*= )*8= +1:== )*= ,T3*8_= *8= 47(-.)S*8= 2>(4-S9S749745-*8= 7S543)*39= .3).7*(9*2*39= A= (*99*= 2U2*= 6:*89.43= *3= 8*= 34:77.88&39= )*= (&7'43*= 574):.9= 5&7= )*8= &7'7*8= 6:*11*8= 7*O4.;*39= )*= (-&25.,3438= *(942>(47-.?.*38= (422:38_= S&324.38`= (*8= 47(-.)S*8= 8*= (425479*39= 9*1= )*8= 5&7&8.9*8= 8:7= (*8=(-&25.,3438=*(942>(47-.?.*38`=*9=)S94:73*39=1*=+43(9.433*2*39=3472&1=)*=1&=8>2'.48*= *(942>(47-.?.*33*`= 1*:7= S9:)*= 3*= )S24397*= 5&8= ).7*(9*2*39= 1&= 89&'.1.9S= )*= 1&= 897&9S,.*= ,S3S7&1.89*=(-*?=(*8=(-&25.,3438_= *9=&79.(1*=(425&7*=1&=).++S7*3(.&9.43=5&7=1&=).89&3(*=A=1&=).++S7*3(.&9.43=8*143=1-9*=a= 8.1=>=&=)*=1&=85S(.&1.8&9.43`=43=8&99*3)=A=(*=6:*= )*8= 545:1&9.438= S14.,3S*8= 84:8= 1&= 2U2*= *85T(*=)&7'7*=84.*39=51:8=574(-*8=*397*=*11*8=6:*=)*8=545:1&9.438=574(-*8=84:8=)*8=*85T(*8= )&7'7*8=).++S7*39*8_= 497*=S9:)*=24397*=6:*= &((&7.&=&2*9->89.3& `=:3=(-&25.,343=)*=+47U9=2&9:7*=&88*?= (422:3`=*89=343=8*:1*2*39=57S8*39=84:8=)*=342'7*:==-9*8=&7'7*8`=2&.8=3*=57S8*39*=5&8= )*= ).++S7*3(.&9.43= 5&7= 1&= ).89&3(*= 3.= 5&7= 1-9*= *9= )43(= 3*= 9*3)= 5&8= A= 8*= 85S(.&1.8*7`= (*= 6:.= )S24397*=1&=89&'.1.9S=)*=(*=(&7&(9T7*=,S3S7&1.89*=)&38=(*99*=*85T(*_= Molecular Ecology (2008) doi: 10.1111/j.1365-294X.2008.03790.x

EvidenceBlackwell Publishing Ltd from population genetics that the ectomycorrhizal basidiomycete Laccaria amethystina is an actual multihost symbiont

MELANIE ROY,* MARIE-PIERRE DUBOIS,* MAGALI PROFFIT,* LUCIE VINCENOT,* ERICK DESMARAIS† & MARC-ANDRE SELOSSE* *Centre d’Ecologie Fonctionnelle et Evolutive (CNRS, UMR 5175), Département Biologie des Populations, Equipe Interactions Biotiques 1919 Route de Mende, 34293 Montpellier Cedex 5, France, †Institut des Sciences de l’Evolution (ISEM, UMR 5554), Equipe Biologie Intégrative, Place Eugène Bataillon, 34095 Montpellier Cedex 05, France

Abstract It is commonly assumed that ectomycorrhizal (ECM) fungi associated with temperate forest tree roots are not host-specific. Because this assumption relies on species delineations based on fruitbodies morphology or ribosomal DNA sequences, host-specific, cryptic biological species cannot be ruled out. To demonstrate that Laccaria amethystina has true generalist abilities, we sampled 510 fruitbodies on three French sites situated 150–450 km away from each other. At each site, populations from monospecific stands (Abies alba, Castanea europea and Fagus sylvatica) or mixed stands (F. sylvatica + Quercus robur or Q. robur + Carpinus betulus) were sampled. Three different sets of markers were used for genotyping: (i) five microsatellite loci plus the ribosomal DNA intergenic spacer, (ii) the mitochondrial large ribosomal DNA subunit, and (iii) direct amplification of length polymorphism (DALP), a new method for fungi providing dominant markers. Evidence for allogamous populations (with possible inbreeding at local scale) and possibly for biparental mitochondrial inheritance was found. All markers congruently demonstrated that L. amethystina popu-lations show little structure at this geographical scale, indicating high gene flow (as many as 50% of founding spores in all populations being of external origin). Our results also showed that host species contributed even less to population differentiation, and there was no evidence for cryptic biological species. This first in situ demonstration of a true multihost ability in an ECM species is discussed in terms of ecology and evolutionary biology. Keywords: ectomycorrhizal fungi, gene flow, host specificity, microsatellite loci, ribosomal DNA, temperate forest Received 7 December 2007; revision received 24 February 2008; accepted 7 April 2008

Introduction 2006). Although specialist (single-host) ECM taxa exist, such as in Leccinum (den Bakker et al. 2004) or Alnicola In temperate forests, tree roots associate with soil fungi to (Moreau et al. 2006), multihost ECM fungi comprise form the ectomycorrhizal (ECM) symbiosis, which has a between 12% and 90% of ECM fungal communities crucial contribution to the nutrition of both partners and (Kennedy et al. 2003; Richard et al. 2005; Twieg et al. 2007). to forest ecosystem functioning (Smith & Read 1997). The The ecological significance of this is that even if multi- ECM symbiosis is considered nonspecific, since most ECM host fungi are present at low frequencies, the connection fungi colonize several host species and are thus generalists of roots of different plants by multihost mycorrhizal (‘multihost fungi’, Bruns et al. 2002). Reciprocally, ECM fungi might drastically affect plant communities (Selosse plants associate with various fungal species (Selosse et al. et al. 2006), e.g. by providing pathways for interspecific transfers of carbon (Tedersoo et al. 2007) or mediating Correspondence: M.-A. Selosse. Fax: (33) (0)467 412138; competitive interactions between plants (Bever 2002). E-mail: [email protected] Multihost ECM fungi may contribute to ecological

© 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd 2 M. ROY ET AL . succession, as in Mediterranean ecosystems, where shrubs geophilum (Douhan & Rizzo 2005), Tricholoma scalpturatum associated with ECM fungi likely facilitate the estab- (Gryta et al. 2006) or Amanita muscaria (Geml et al. 2006). lishment of ECM trees, by functioning as inoculum reservoirs CBSs seem to occur in most investigated morphospecies, (Richard et al. 2005). even if this may result from underreporting of models However, evidence for multihost ECM fungi relies on lacking cryptic speciation. Arguably, fungal species have questionable definitions of the fungal species. A classical fewer morphological characters than macro-organisms, approach is based on a morphological definition (morpho- and speciation could less impact morphology (Taylor et al. species). The occurrence of fruitbodies (the sporulating organ 2006). Thus, closer investigation of supposedly multihost formed by many ECM fungi) with identical morphology ECM fungi is required to rule out the possibility of host- under various host trees is considered as an evidence that specific CBSs — and this is the aim of the present work. the species is multihost. More recently, molecular approaches Analysis of gene flow in natural populations can assess assume that identical DNA sequences (e.g. the internal whether a supposed ‘species’ is a single breeding unit, or is transcribed rDNA spacer, ITS) equate to identical species. divided into several CBSs (Grünig et al. 2007), circumventing Indeed, there is good congruence between ITS sequence mating tests for uncultivable strains. We focused on Laccaria and morphospecies (Horton 2002), but both approaches amethystina, an uncultivable, common species from Eura- potentially overlook cryptic host-specificity. Phylogenet- sian forests, belonging to a genus for which host specificity ically related but genetically isolated species that would be seems rare (Kropp & Mueller 1999). Two previous in- host-specific could retain similar ancestral fruitbody vestigations of beech forest populations suggested that morphology and ITS sequence, while preventing any gene L. amethystina populations largely outcross, allowing for gene flow, leading to cryptic biological species (CBS). Inter- flow between populations (Gherbi et al. 1999; Fiore-Donno estingly, in the Leccinum genus, where species differ by & Martin 2001). Here, we test whether L. amethystina host specificity, interspecific morphological differences populations under different host trees (including deciduous are limited (den Bakker et al. 2004). Similarly, ITS diver- and coniferous species) show genetic differentiation by gence might not fully correlate with biological species, if using two independent sets of nuclear markers and a mito- sequences diverge only after genetic isolation, as suggested chondrial marker. Since isolation by distance might generate in ECM Hebeloma spp. (Aanen et al. 2000). Moreover, direct genetic variation between populations independently of evidence that an ECM fungal genet simultaneously associates the host (Bergemann & Miller 2002), and since no data are with two host species is so far limited to (i) ex situ inocula- currently available at scales above 50 m, we also investi- tion experiments, where experimental conditions might gated population genetic structure at two scale magni- allow artefactual associations (Selosse et al. 2006), and (ii) tudes, namely 1 km and 100 km. By analysing populations rare in situ genotypings based on a single locus (Taylor & from three different forest sites in France, we tested whether Bruns 1997; Selosse et al. 2002). Although multihost ability host tree species contribute more to population genetic in ECM fungi is probably the rule, it still awaits rigorous structure than geographical distance. demonstration. Species definition and speciation are debated for fungi Materials and methods (Kohn 2005; Taylor et al. 2006), but CBSs are often described within morphospecies. Lack of recombination can be dem- Model species onstrated by (i) in vitro mating tests between haploids, (ii) study of gene flow, or (iii) analyses of multigene phyloge- Laccaria amethystina (Cooke) is a typical basidiomycetous nies (phylogenetic species recognition; Taylor et al. 2000). species, where the dikaryotic mycelium (diploid thallus Ecological specialization is a driving force in sympatric arisen from mating, whose cells harbour pairs of different fungal speciation (Kohn 2005), and host specialization haploid nuclei) grows vegetatively. It colonizes roots correlates with CBSs in many parasitic taxa, such as the and forms above-ground fruitbodies bearing meiotic spores anther smut fungus Microbotryum violaceum on Caryophyl- (Moore & Novak Fraser 2002). Establishment of these laceae species (Le Gac et al. 2007), or the root rot fungus haploid spores, followed by mating, creates new dikaryotic Heterobasidion annosum on spruce, pine and silver fir mycelia. Dikaryotic genets can thus be typed by way of (Gonthier et al. 2001). Host specificity could also drive fruitbody analysis. speciation among ECM basidiomycetes, as proposed in the Hebeloma crustuliniforme species complex (Aanen et al. 2000), Sampling and DNA extraction among Xerocomus spp. (Taylor et al. 2007), or among the suilloids (Kretzer et al. 1996). Moreover, even without rela- A total of 510 fruitbodies of L. amethystina were collected in tionship to host preference, CBSs are frequently reported fall 2001 from three different French forests situated from among ECM fungi, such as Pisolithus tinctorius (Martin et al. 150 to 450 km away from each other (Fig. 1a). In each forest, 2002), Cantharellus formosus (Dunham et al. 2006), Cenococcum we chose stands with only one tree species (Abies alba,

were stored at –20 °C within 3 h of collection. Eight French dried fruitbodies of other Laccaria spp. provided by P.-A. Moreau were used as outgroups in our analyses: L. laccata var. pseudobicolor [PAM01042909], L. laccata var. moelleri [PAM97090101], L. laccata var. pallidifolia [PAM01043006], L. oblongispora [PAM01042805], L. macrocystidiata [PAM99082801 and PAM00103002] and L. proxima [PAM01102404 and PAM01110104] (numbers are accessions in P.-A. Moreau herbarium at Université de Lille). For DNA extraction, 100 mg fruitbody pieces were ground in 1.5-mL Eppendorf tubes using 1.1-mm diameter Tungsten carbide balls (Biospec Products) in a Retch MM301 vortexer at 30 Hz for 2.5 min. Extraction was performed using the DNeasy Plant Mini Kit (QIAGEN) according to the manufacturer’s instructions. DNA was recovered in 100 µL distilled water. Sequencing of ITS of 70 randomly chosen fruitbodies provided five different sequence types (GenBank Accession nos EU076450–EU076454), diverging by only 1- to 2-point mutations and uncorrelated to host or forest of origin.

Microsatellites and IGS markers Microsatellites were characterized as in Sarthou et al. (2003), using a genomic library derived from a pool of two Orry fruitbodies. After screening of about 3000 colonies

using the oligonucleotides (TC)10, (TG)10, (CAC)5CA, CT(CCT)5, CT(ATCT)6 and (TGTA)6TG, a total of 34 positive clones were sequenced. Microsatellite repeats were present in 17 loci (EF444487–444503). Flanking primer pairs were designed using the primer 3 software Fig. 1 The investigated French populations. (a) location of the (http://froda.wi.mit.edu/cgibin/primer3/primer3_www.cgi; three investigated French forests, with FST values and distances Sarthou et al. 2003), and tested on 45 randomly selected between them. (b) F values between populations from different ST samples from the three investigated forests (15 from each host trees and from different forest. On the right, comparison forest) for repeatability and polymorphism of patterns. of means (± standard deviation) of FST calculated between populations pooled from the same host tree (grey) and from the Only La171 (EF444487) and La115 (EF444488) fitted these same forest (black). On the left, comparison of means (± standard criteria (Table 2). These samples were used to screen deviation) of FST calculated between pairs of populations differing 10 loci obtained by Wadud et al. (2006) from Japanese L. by host (grey) or by forest of origin (black). amethystina and seven loci obtained by Jany et al. (2006) from American L. bicolor. We, respectively, selected the markers La03, La06, La17, La21 and La23 from Wadud et al. (2006) and LBTC38 from Jany et al. (2006) (Table 2). Castanea europaea or Fagus sylvatica) or a mix of two We also used the intergenic spacer 1 (IGS1) of the rDNA species exclusively (Fagus sylvatica + Quercus robur or that provides an amplicon polymorphic in size when using Q. robur + Carpinus betulus), where we collected our popu- primers CNL12 + 5SA (Selosse et al. 1996) and behaves as lations. The 13 resulting populations, distributed across a Mendelian locus (Selosse et al. 1996, 1998). In order to the three forests, were located at less than 2.5 km away multiplex it with microsatellites, the amplicon size was from the other in each forest (minimal distance: 0.5 km; see reduced by designing a primer internal to IGS1. Briefly, we Table 1 for the populations’ names). Since average and found a conserved IGS1 region in which the primer IGA maximal genet diameter found in previous studies were was designed (Fig. S1a, Supplementary material; Table 2). 0.65 m and 1.5 m (Gherbi et al. 1999) and 1.1 m and 5.4 m The fragment amplified using IGA and 5SA, flanking the (Fiore-Donno & Martin 2001), respectively, we collected 5.8S rDNA, contains all the IGS1 size polymorphism [data fruitbodies situated at more than 1.5 m from each other to not shown; polymerase chain reaction (PCR) conditions as limit redundant sampling of the same genet. Fruitbodies in Selosse et al. 1996].

Table 1 Features, names and locations of the Laccaria amethystina populations under study. Populations were named with a first letter to identify the forest (P, B or O) followed by letter(s) identifying the host trees (f, a, c, fq or bq)

Forest name Pilat (P) Bellême (B) Orry (O) Altitude (above sea level) 1000 m 130 m 40 m Geographical position 45°21′N, 4°29′E 48°23′N, 0°31′E 49°07′N, 2°28′E

Fagus sylvatica stands (f) Pf Bf Of No. of fruitbodies 61 30 44 Area size (m2) 240 600 400 Minimum age of the trees (years) 150 170 150 Origin of the trees NR NR NR Abies alba stands (a) Pa Ba Oa No. of fruitbodies 43 68 5 Area size (m2) 200 1000 200 Minimum age of the trees (years) 80 55 30 Origin of the trees NR NR P

Castanea europaea stands (c) Pc Bc Oc No. of fruitbodies 46 44 47 Area size (m2) 300 200 200 Minimum age of the trees (years) 50 50 100 Origin of the trees NR NR NR

Fagus sylvatica + Quercus robur stands (fq) Not found Bfq Ofq No. of fruitbodies 32 8 Area size (m2) 600 50 Minimum age of the trees (years) 50 150 Origin of the trees NR NR

Carpinus betulus + Quercus robur stands (bq) Not found Bbq Obq No. of fruitbodies 46 36 Area size (m2) 2000 2025 Minimum age of the trees (years) 15 50 Origin of the trees NR NR

*Stands are naturally regenerated, with local trees (NR), or planted (P) with trees from nurseries.

For genotyping, microsatellite loci and the polymorphic by sequencing amplicons from primers ML5 and ML6 (as IGS1 fragment were amplified using the PCR Multiplex kit in Selosse et al. 1998) on the 45 L. amethystina samples (QIAGEN) protocol. Reactions were performed on a previously selected. The PCR thermoprofile included: PTC-100 programmable Thermo Controller (MJ Research) initial denaturation at 94 °C for 4 min, followed by 35 under the following thermoprofile: initial denaturation at cycles of denaturation at 94 °C for 30 s, annealing at 50 °C 95 °C for 15 min, followed by 30 cycles of denaturation for 30 s and extension at 72 °C for 30 s, with a final at 94 °C for 30 s, annealing at 58 °C for 90 s and extension extension at 72 °C for 30 s. Two polymorphic sites were at 72 °C for 60 s, with a final extension at 60 °C for 30 s. Forward detected, namely at positions 60 (C or A) and 120 (T or G) primers were labelled with different fluorochromes. Detec- downstream of the 5′ end of ML5 (Fig. S1b). Only three tion of labelled PCR products was carried out on an ABI haplotypes were recovered: C60T120 (haplotype α; GenBank PRISM 3130 XL Genetic analyser (Applied Biosystems). Accession no. EF444506), A60G120 (haplotype β; GenBank For this, a 3-µL aliquot from a 1:1000 dilution of the PCR Accession no. EF444507) and A60T120 (haplotype γ; mixture was pooled with 15 µL of Hi-Di formamide and GenBank Accession no. EF444508). The primer AML5.5 0.2 µL of standard Genescan 500 ROX size ladder (Applied (5′-TATAAGCTAATTTATTTATATTC-3′) was designed Biosystems). Fragment sizes were analysed with gene- to detect haplotype α when paired with ML5.5 (5′- mapper 3.7 (Applied Biosystems), using default parame- AAACACAGTGGCAATATTCAAT-3′), since its 3′ end ters for microsatellite analysis. matches the specific C60 base of this haplotype. The primer AML5 (5′-TGTGTCACCGTTATAAGTTC-3′) detected haplotype β when paired with ML5, since its 3′ end matches Mitochondrial genotyping the specific base (G120) of this haplotype (see Fig. S1b Polymorphism of the large subunit of mitochondrial for primer positions; PCR conditions as above, except for ribosomal DNA (LrDNA) was preliminarily investigated annealing temperature: 56 °C). The haplotype was considered

Table 2 Nuclear markers (microsatellites, IGS and DALPs) and related primers used in this study

Marker Primers sequence (5′–3′) Tm (°C) Fragment size (bp) Allele no. Reference

Microsatellite La171* CGCTCAAAACTGCACCAAC 56 263–297 12 This study TGTTCCCTAATCAACATACCC La115† GCAGGAGAGTGAACCATGTG 60 123–140 11 This study ACCCCAAACTCAAAGCATTC La03 GAGAAAAATTGGTGAAACCCA 56 150–157 6 Wadud et al. 2006 CAACTATGACATGCCAATTCG La17 GCTGGGTCTCTCTTCCTAATC 56 119–137 7 Wadud et al. 2006 GGAGTTGCGAAAGAGACATAG La23 GCCATCGTCGGATCAGCTTAC 56 227–239 6 Wadud et al. 2006 AGATTGAGAAGATGTACGAG LBTC38 CGCACGAGTCTGATAACGAG 56 116–130 7 Jany et al. 2006 TTCCGTCACCGTAGTCACAG La06 GAATCACAAACCCACAGAATC 56 183–186 2 Wadud et al. 2006 CTTTCGTCGACCCGAATTATG La21 CTGGCTGTTTCGCTGTATAGT 56 152–155 2 Wadud et al. 2006 GAAGTAGATGTCACACTGGATG Nuclear rDNA IGS‡ IGA: CATTTTGACTTGCSATTGAGG 55 237–309 9 This study 5SA: CAGAGTCCTATGGCCGTGGAT DALPs§ 232 GTTTTCCCAGTCACGACGAC 50 100–500 12 Desmarais et al. 1998 233 GTTTTCCCAGTCACGACACG 50 100–500 21 Desmarais et al. 1998 235 GTTTTCCCAGTCACGACCAC 50 100–500 11 Desmarais et al. 1998

Tm, annealing temperature. *repeated motif: (GAG)7 ... (GAG)3 (TG)3. †repeated motif: (CAA)3(CA)7 (CT) 3 (CAA)4. ‡see Fig. S1a for primer position. §reverse primer is 5′-TTTCACACAGGAAACAGCTATGAC-3′ for all DALPs (Desmarais et al. 1998).

to be γ whenever no amplicon was amplified using AML5 vidual for electrophoresis separation on an ABI PRISM 3130 XL + ML5 and AML5.5 + ML5.5, although a PCR product was Genetic analyser as described for microsatellites. For each obtained using the ML5 + ML5.5 primer set (positive PCR PCR, a control with sterile water instead of DNA was per- control; all PCRs were repeated twice). This primer design formed. DALP fingerprints were analysed with genemapper allowed congruent identification of the LrDNA haplotype 3.7 (Applied Biosystems), using AFLP analysis method on the 45 preliminary samples. PCR products were checked with default parameters. We excluded fragments that were on a 0.8% agarose gel stained with ethidium bromide nonfully reproducible or shorter than 100 bp, and all finger- (0.5 µg/µL). prints were checked by eye twice. Due to possible variations of migration time on ABI PRISM, fragments differing by ± 1 bp were considered identical. DALP genotyping Direct amplification of length polymorphism (DALP), a Data analysis method providing polymorphic, dominant fingerprints (Desmarais et al. 1998), was used with three different Populations Oa and Ofq were retrieved from the analyses primers (DALP232, DALP233 and DALP235) that revealed (except for DALPs) because of their small sizes. In order to polymorphisms in preliminary tests on the 45 L. calculate the allelic frequencies, departure from Hardy– amethystina samples previously selected (not shown). A Weinberg equilibrium (Table 3), and linkage between loci, subsample combining half of each population and all we used genepop′007 (Rousset 2008). A correspondence individuals from populations Oa and Ofq (370 individuals factorial analysis on populations using microsatellite in all) was used for DALP genotyping. Reproducibility data was performed with genetix 4.05.2 (Belkhir et al. of DALP fingerprints was tested by replicating DNA 1996–2004) to detect any Wahlund effect. Population extractions on all individuals. Dye-labelled primers were differentiation pairwise tests, Wright indices (FIS, FST and used to allow detection. Amplifications were performed FIT) and Slatkin’s indices (RST) were estimated using genepop separately for each primer, according to conditions of ′007. Correlation between genetic distances (FST or Desmarais et al. (1998). Amplicons were pooled per indi- RST) and geographical distances in an island model was

Table 3 Summary of population analysis by the five polymorphic microsatellite loci, the IGS and the mitochondrial LrDNA. FIS were calculated using clone-corrected data (fruitbodies from the same population having identical genotypes were taken into account only once).

Level of statistical support: *P < 0.1; **P < 0.01; ***P < 0.001. The selfing coefficient s was calculated as s = 2FIS/(1 + FIS), from clone-corrected data, excluding La171 and La115 that have null alleles

Selfing LrDNA haplotypes Population Identical fruitbodies† La171 La115 La17 La23 L 03 All micr. IGS coefficient s (α/β/γ) in %

Bc 2 × [2] 0.17* 0.08 0.38*** −0.07 1.00** 0.12 −0.02 0.32 52/33/14 Bbq 3 × [2] 0.4*** 0.27* 0.49** 0.05 0.13* 0.27 0.04* 0.48 42/41/17‡ Bf 2 × [2] 0.29** 0.63*** 0.34 0.17 −0.04 0.21 −0.10 0.40 41/36/23 Bfq [2] 0.29** 0.88*** 0.32*** 0.13 0.29 0.26 −0.06 0.41 48/32/19 Ba [3] + 6 × [2] 0.36*** 0.52*** 0.23*** 0.29 0.87** 0.36 0.11* 0.59 59/3/11‡ Oc [3] + [4] + [17] 0.27** 0.13* 0.23** −0.11 1.00** 0.06 −0.03 0.35 46/34/20‡ Obq [2] 0.43*** 0.74*** 0.65*** 0.09 0.66* 0.38 0.02 0.60 53/34/13 Of [2] 0.43*** 0.29* 0.19** −0.17 −0.01 0.17 0.10 0.27 49/31/21 Pc [2] 0.55*** 0.46*** 0.31* −0.06 −0.27* 0.16 0.02* 0.27 43/39/18 Pf None 0.41*** 0.27*** 0.25** 0.03 0.17* 0.23 0.15*** 0.38 49/36/15 Pa 5 × [2] + 2 × [4] 0.20** 0.20* 0.48* −0.04 0.29 0.09 −0.09 0.33 54/22/24 All pop. 0.36 0.38 0.35 0.30 0.35 0.27 0.02 0.27 50/33/17

†identical fruitbodies are fruitbodies from a population sharing identical microsatellite and IGS polymorphism. In the table, numbers of identical fruitbodies are inside brackets (preceded by number of groups with the same number of fruitbodies). ‡heterozygous individual with α and γ were detected in these populations (one each in Bfq and Oc, and two in Bbq).

calculated by a Mantel test implemented in genepop′007. Given allelic frequencies (not shown), and assuming genetic Analysis of molecular variance (amova) was performed independence between markers, the expected likelihood on microsatellite markers, IGS and DALP data set, using of these genotypes ranged from P = 0.031 to P = 0.011 arlequin 2.0 (Excoffier et al. 2000), in order to test whe- (0.0052 for the 17 Oc fruitbodies) when all fruitbodies ther the genetic variation was better explained by the were from the same population, and lower (from 2.2 10–5 to geographical distance or by the host tree. To visualize the 7.5 10–5) when fruitbodies occurred in different populations. genetic structure obtained with DALP, we calculated Dice Although such resemblances might occur by chance, distances between individual DALP profiles with darwin mainly within populations, we cannot rule out that some 4.0.212 (Perrier et al. 2003). The Dice dissimilarity, D, fruitbodies represent the same genet. We used two dif- between two DALP profiles is calculated as D = (b + c)/ ferent data sets in further analyses: one incorporating all (2a + b + c), where a is the number of common bands, b the fruitbodies (raw data) and the second excluding geno- number of specific bands of the first individual and c types repeated within a population to correct for potential the number of specific bands of the second one. Then, clones. the neighbour-joining method (Saitou & Nei 1987) was used For La171, La115 and LBTC38, failure to obtain PCR to build trees from Dice distance matrix, and treedyn products on some individuals suggested the occurrence of (Chevenet et al. 2006) was used to develop a graphic display. null alleles (respective estimated frequencies: 0.05, 0.03, 0.6). Null allele frequencies were higher in all populations for LBTC38 than for other markers, perhaps because it was Results initially designed for Laccaria bicolor. For La06 and La21, one of their two alleles (Table 2) was very rare. La06, Microsatellite and IGS polymorphism La21 and LBTC38 were thus discarded from the analysis. In the 11 populations under study (497 fruitbodies), the The only linkage disequilibrium was found between La23 eight microsatellite loci showed two to 12 alleles per locus and La17 (P < 0.001 using the clone-corrected data set) and (Table 2). IGS was also polymorphic, with nine alleles only four private alleles were detected (two in Bfq and one among which two had high frequencies (0.47 and 0.38). in Pf and Bbq; one fruitbody only in each case). Based on these nine loci, 28 genotypes were found in more than one fruitbody (Table 3). These genotypes occurred Population analyses using microsatellite and IGS loci in two to four fruitbodies in all case, but one: 17 fruitbodies of population Oc were identical. In three occurrences Whenever significant departures from Hardy–Weinberg only, these fruitbodies were from separate populations. equilibrium were observed, each of the five microsatellite

Table 4 FST values for the five polymorphic microsatellite loci plus the IGS among the 11 Laccaria amethystina populations under study (in bold, FST significant at P < 0.01; boxed FST values are these of population pairs growing under the same host species). FST in the upper triangle are calculated using all fruitbodies, and FST in the lower triangle are calculated with clone corrected data

Bc Bbq Bf Bfq Ba Oc Obq Of Pc Pf Pa

Bc 0.010 0.246 −0.004 0.078 0.075 0.005 0.019 0.032 0.008 0.019 Bbq 0.009 0.206 −0.0002 0.059 0.057 0.006 0.018 0.037 0.018 0.006 Bf 0.243 0.203 0.202 0.060 0.259 0.235 0.236 0.139 0.206 0.215 Bfq −0.005 −0.001 0.199 0.044 0.059 −0.002 0.016 0.027 0.013 0.015 Ba 0.073 0.054 0.059 0.041 0.118 0.077 0.084 0.035 0.072 0.076 Oc 0.021 0.004 0.215 0.009 0.068 0.044 0.055 0.104 0.080 0.067 Obq 0.005 0.005 0.232 −0.004 0.073 0.001 0.023 0.046 0.029 0.022 Of 0.021 0.018 0.237 0.014 0.083 0.009 0.023 0.028 0.015 0.008 Pc 0.031 0.035 0.138 0.024 0.031 0.044 0.043 0.029 0.011 0.023 Pf 0.007 0.017 0.204 0.018 0.066 0.026 0.026 0.018 0.011 0.005 Pa 0.014 0.006 0.210 0.011 0.066 0.004 0.015 0.004 0.015 −0.0002

loci showed significant heterozygote deficiencies (Table 3) by distance (R2 = 0.0147, P = 0.63, Fig. S3, Supplementary in all populations, except La03 in population Pc that had material). Moreover, an average of 11.4 migrants per popu- significant heterozygote excess. (The too-small Oa and Ofq lation was calculated, i.e. about 29% of each population. FST populations were discarded from the analysis.) Using values of population pairs growing under the same host clone correction or raw data did not modify these results species were low (ranging from 0.0045 to 0.23; FST value (not shown), and the whole 497 fruitbody pool also showed boxed in Table 4), and mean FST of such pairs (0.08) was heterozygote deficiencies (Table 3). The IGS locus ex- slightly higher than the mean value of random population hibited heterozygote deficiencies in five populations (four pairs (0.05). As these FST values were low, we pooled indi- were significant, both on raw and corrected data), and viduals growing in the same forest: FST among forests nonsignificant heterozygote excess in the seven other ranged from 0.0181 to 0.0293 (all nonsignificant, Fig. 1a). (significant in Pa on raw data only; Table 3). Heterozygote Similarly, we pooled populations growing under the same deficiencies could be explained by a Wahlund effect, but host: FST ranged from 0.0001 to 0.0299 (all nonsignificant; a correspondence factorial analysis failed to split the Fig. 1b). FST between different hosts were lower than FST samples into distinct populations (the three first axes between regions (Fig. 1): although this difference was not explained only 15% of the variability; Fig. S2, Supplementary significant (P = 0.11 according to Mann–Whitney test), this material). Thus, genetically divergent subpopulations supported the hypothesis that host species did not drive were unlikely to have been sampled within each of population differentiation more than distance. All these our populations. Similarly, null alleles could lead to analyses gave identical trends using RST values (not overestimation of the homozygosity level, but markers that shown), but the sequence of some markers (IGS, La115 and lack detectable null alleles had positive FIS values (Table 3). La171, Table 2) did not fit with the theoretical criterions for Assuming that autogamy accounts for heterozygote applying RST calculations. deficiencies, and omitting La171 and La115 whose null Furthermore, 0.01% of the variance in an amova was alleles would bias the values, selfing coefficients ranged explained by the forest (i.e. the geographical origin), from 0.09 to 0.55 among populations (0.35 for the whole 497 whereas only –0.04% of it was explained by the host species fruitbodies) on raw data, and from 0.27 to 0.60 (Table 3) (Table 5). Moreover, genetic variance among populations after clone correction. within a forest was lower (0.09%) than that among popula- Pairwise population differentiation tests supported the tions growing under the same host species (0.13%; Table 5). hypothesis that all populations differed significantly (not Thus, groups of populations were more diverse when shown). Most FST values between populations were posi- grouped by host-species than when grouped by forest, and tive, but low and not significant, except for populations Bf, no clear differentiation was explained by host species.

Ba and Oc whose FST were both high and significantly supported (Table 4). These results were not modified by taking Population analysis by mitochondrial LrDNA null alleles into account or by ignoring loci with null alleles polymorphism (not shown), nor after clone correction (except for significance of FST for Oc, Table 4). However, a Mantel test The three mitochondrial haplotypes occurred in each revealed no significant differentiation among populations population (Table 3). Four individuals cumulated α and γ

Table 5 Distribution of microsatellites and Populations IGS alleles or DALP fingerprints covariance, pooled by: Microsatellite + IGS DALP according to an amova, within and between groups of populations pooled either by Forest Among forests 0.01% 0.02% forests (= geographical origin) or by host Among populations within forests 0.09% 0.19% trees Within populations 1.59% 2.12% Host tree Among host pool −0.04% 0.01% Among populations within host pool 0.13% 0.19% Within populations 1.59% 2.12% Both Total 1.68% 2.32%

Mean FST over all loci 0.06% 0.08%

types (Table 3): repetition of DNA extraction and LrDNA for under-dispersed fruitbodies. According to an amova, sequencing showed that they were heteroplasmic at 0.02% of the variance of DALP fingerprints was explained position 60 (not shown). In all populations, the α haplo- by the forest (= geographical origin), while 0.01% was type was the most frequent (0.59), while γ was the rarest explained by the host species (Table 5). Both genetic vari- (0.11). Among the 28 groups of fruitbodies having iden- ances among populations within a forest and among tical microsatellite and IGS genotype, seven exhibited two populations growing under the same host species were different mitochondrial haplotypes, depending on the low (Table 5). Congruently with microsatellite data, the fruitbody, and three exhibited three different haplotypes, host species factor is thus less explanatory than the geo- suggesting that these fruitbodies were not necessarily from graphical one. the same genet. The frequencies of the three types did not differ significantly among populations (χ2 = 23.658, d.f. = 20, Discussion P = 0.258) and among the three forests (χ2 = 0.179, d.f. = 4, P = 0.996). The haplotype frequencies obtained by pooling Based on three polymorphic marker sets (five micro- fruitbodies from the same host species did not differ satellite loci plus the IGS, a mitochondrial gene, and DALP among the different hosts (χ2 = 4.136, d.f. = 8, P = 0.845). fingerprints), Laccaria amethystina populations showed little structure over c. 500 km (Fig. 1). Host tree species did not detectably contribute to population genetic structure Population analysis by DALP markers (Table 5), thus making L. amethystina a true multihost DALP patterns were highly polymorphic and, from the 370 biological species. This is, to our knowledge, the first investigated individuals, 44 fragments were reproducible demonstration of a multihost ability in ECM fungi by and informative (i.e. present on more than one individual). population analysis, although it is likely to be valid in Based on frequencies of DALP fragments, expected many species. We also report on the reproductive biology likelihood of DALP fingerprints ranged from 0.01 to 0.14. of this species (i.e. a trend toward heterozygote deficiency and Twenty groups of fruitbodies with identical microsatellite indirect evidence for biparental mitochondrial inheritance). and IGS genotype (Table 3) could be distinguished by DALP fingerprints, used here for the first time on a fungal DALPs, showing the fruitbodies were unlikely to belong to model, produced easy and fast polymorphic patterns of the same clone. However, in the eight remaining groups dominant markers. Interestingly, amova on DALPs and of fruitbodies, no difference in DALP fingerprint was microsatellites were congruent (Table 5). In this species for seen [e.g. among the 17 fruitbodies from population Oc which suitable microsatellites are difficult to find, DALPs previously mentioned (Table 3; see star on Fig. 2)]. High provided useful additional tools to distinguish genets. genetic similarity or clonality might thus explain these Cumulating all markers, the probability of encountering cases. a genotype by chance ranged from 1.16 × 10–13 to 5.02 × 10–3, In a phylogram recapitulation (Fig. 2), other Laccaria spe- well within the range of other studies (e.g. Kretzer et al. cies used as outgroups clustered together and no major 2005; Bergemann et al. 2006; Lian et al. 2006). In all, 34 gen- geographical clustering within L. amethystina was revealed. otypes were found in more than one fruitbody. To avoid Some fruitbodies from the same forest or the same host repetitive sampling of the same genet, fruitbodies were col- tree species clustered together (Fig. 2). However, 84% of lected at more than 1.5 m from each other, i.e. more than the fruitbodies found under the same host tree that clustered average genet diameter previously reported (Gherbi et al. in this analysis were also from the same population, so 1999; Fiore-Donno & Martin 2001). The finding of identical that clustering by geographical origin was the likely reason fruitbodies could result from infrequent large clones, or

Fig. 2 NJ phylogram using Dice distances based on DALP data (grey branches: Laccaria spp. outgroups). The phylogram is represented twice to show congruence between geographical and host clusters: on the left, branches are coloured by forest (yellow, Bellême forest; red, Orry forest; blue, Pilat forest); on the right, colours represent host trees (yellow, Castanea europea; red, Quercus robur + Carpinus betulus; green, Fagus sylvatica; black, Fagus sylvatica + Quercus robur; blue, Abies alba). The star indicates the large cluster of 17 fruitbodies from Oc population (see text).

from an underestimation of genet size in previous studies, was reported when crossing haploids with divergent because the RAPD markers used were sensitive to contam- LrDNAs (Selosse et al. 1998), while most siblings were inants. Genetically similar genets could also occur in close homoplasmic, probably as a result of mitochondrial vicinity. Interestingly, the 17 fruitbodies from population sorting during zygotic growth. The frequency of α + γ Oc (Table 3; star on Fig. 2) with identical DALP, IGS and heteroplasmic genets (0.006) is far lower than expected microsatellite fingerprints, showed three different LrDNA under random mating (0.057): under the assumption of haplotypes, suggesting that they arose from different biparental inheritance, mitochondrial sorting and/or a haploid parents (see below). Moreover, for this cluster, the trend to inbreeding (Table 3) could explain this. The haplo- –3 probability of occurrence by chance (2.65 × 10 ) is much type C60G120 could be absent because of (i) a lack of lower than the observed frequency (0.38), further supporting recombination between haplotypes α and β, and/or (ii) genetic relatedness. However, since clone correction did a counter-selection of this sequence for proper ribo- not change our results, the exact status of these fruitbodies somal functioning, whenever mutation or recombination did not influence our conclusions. produce it. The later possibility is a limitation encountered when using sequences under selection; however, since this selection is unlikely to depend on host trees, it does not Mating system in L. amethystina modify our conclusions on nonspecificity of L. amethystina. The mitochondrial LrDNA markers showed (i) the Nuclear markers showed a trend to heterozygote existence of rare heteroplasmic α + γ individuals (α + β and deficiencies in all investigated populations (Table 3). This β + γ were not detectable in our design), and (ii) the absence was neither explained by fruitbodies with identical geno- of the fourth possible haplotype, namely C60G120 (detectable type (clone correction did not modify the results), nor by in our design). Heteroplasmies arose either by mutation in null alleles, since microsatellites without detectable null a α- or γ-only background, or after biparental inheritance. alleles also had positive FIS (Table 3). It could result from Although mitochondrial inheritance is variable in fungi sibling mating and selfing, due to the heavy self-spore rain (Xu et al. 2005), symmetrical cytoplasm mixing during under fruitbodies on existing genets. Although two mating- mating allows biparental inheritance in basidiomycetes. In type loci exist in Laccaria spp. (Kropp & Mueller 1999), each the related Laccaria bicolor, a low frequency of heteroplasmy haploid can mate with 25% of the other haploids arising

© 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd 10 M. ROY ET AL . from the same dikaryon (Moore & Novak Fraser 2002). In species with epigeous fruitbodies disperse by wind, likely addition, a dikaryon can provide a compatible haploid over long distances. The spore bank is less important and, nucleus to 50% of its haploid progeny: the so-called ‘Buller even if most spores land close to the parental fruitbody, phenomenon’ is described in vitro for L. bicolor (de la long-distance gene flow is possible. As a result, differe- Bastide et al. 1995) and in situ for saprobic basidiomycetes ntiation over 100–1000 m scales is lacking (e.g. no detectable (Johannessona & Stenlid 2004). Thus, inbreeding could structure over 230–1090 m was found in Russula brevipes; lead to an increased homozygosity, and perhaps clumps of Bergemann et al. 2006). There are also intermediate models genetically related fruitbodies, as previously discussed. A such as Suillus grevillei (FST = 0.020 over 700 m; Zhou et al. positive relation between genetic similarity and distance at 2001) or Cantharellus formosus (detectable structure at more metric scale was already reported for several ECM fungi than 400 m only; Dunham et al. 2006). L. amethystina was (Zhou et al. 2001; Liang et al. 2004, 2005; Dunham et al. previously shown to exhibit very moderate structure at

2006); Gryta et al. (2000) described large Hebeloma cylin- small scales (FST ranging up to 0.02 for IGS over 45 m; drosporum genets with highly related individuals in Gherbi et al. 1999). In our study, no structure was found their neighbourhood. Analysis of repartition of related over 450 km (Tables 4 and 5) and FST values did not L. amethystina fruitbodies would further support this correlate with distance (Fig. S3). This is reminiscent of explanation, but our sampling design does not allow wind-dispersed saprophytic and parasitic basidiomycetes such fine-scale spatial analysis. (e.g. Kauserud & Schumacher 2003). The average migrant In previous analyses of L. amethystina IGS polymorph- proportion (29%) relates to dikaryotic genotypes: they ism, there was some discrepancy as to whether heterozy- occurred from mating between two migrant haploid gotes were in excess (Gherbi et al. 1999) or in small deficit spores, whereas genotypes occurring from mating between (Fiore-Donno & Martin 2001) for this locus. First, the highly migrant and local spores might remain undetected. Thus, sensible allele detection by heteroduplex formation used the proportion of migrant spores might be as high as by Gherbi et al. (1999) likely revealed more alleles and 0.291/2 = 0.54. heterozygotes that went undetected in the study by Most populations of L. amethystina showed nonsignifi- Fiore-Donno & Martin (2001). Second, under the previous cant differentiation within forests, and even among forests assumption of a fine-scale structure, plots of the latter (Fig. 1), with the exception of three (namely two Bellême study (5 × 10 m) were smaller than those of Gherbi et al. populations, Bf and Ba, and an Orry population, Oc; (1999; 10 × 10 m) and might be more impacted by local Table 4). Reasons for these highly differentiated popula- inbreeding as described above. Third, in our study, IGS are tions are unclear, but at least unrelated to (i) tree age or an closer to Hardy–Weinberg equilibrium than are microsatellite origin from a nursery, as these populations have different loci (Table 3). IGS might be genetically linked to a mating- ages and are naturally regenerated (Table 1), (ii) the forest type locus, and thus driven to higher heterozygosity levels. of origin, since other populations from these forests are not Selosse et al. (1996) demonstrated that 6.5% of haploids differentiated, or (iii) the host tree, since three different tree in L. bicolor inherit IGS copies from both parents, due to species are involved. a crossing over within the rDNA locus: the occurrence of The 500-km scale was hitherto poorly investigated for such haploids in L. amethystina populations could increase ECM species. Distance of 50 km can lead to some isolation the apparent heterozygosity level. Last, as for LrDNA, (as for T. populinum; Gryta et al. 2006) and distance over selective pressures might bias this marker. This could also 1000 km can lead to stronger differentiations, e.g. for account for (i) the high frequency of two alleles in this and hypogeous species such as Tuber melanosporum (FST = 0.20 other works (Gherbi et al. 1999; Fiore-Donno & Martin between Italy and northern Spain; Murat et al. 2004) or 2001), and (ii) the conserved sequence found within the putatively asexual species such as Cenococcum geophilum

IGS (see Fig. S1a). (FST = 0.25 over North America; LoBuglio et al. 2002). Bergemann & Miller (2002) reported an FST = 0.43 for Russula brevipes over 1500 km in the western USA, but Spatial population genetic structure in L. amethystina the lack of shared alleles between the populations strongly In current studies, ECM fungal population structures questioned their conspecificity. For L. amethystina, the range between two extreme types. ECM species with low distance at which all populations significantly differ is dispersion, and thus high structure over short (kilometric- not reached in this study, and requires samplings at longer scale) distances, are exemplified by hypogeous species distances. L. amethystina is considered to be a Eurasiatic dispersed by animals (Kretzer et al. 2005). Due to their species (Kropp & Mueller 1999; Wadud et al. 2006), but the below-ground fruiting, these species accumulate large low portability of microsatellites from Japanese strains spore banks in soils that outcompete migrants: FST values in our study suggests that some differentiation might exist reached 0.26 over 8.5 km for Rhizopogon occidentalis over Eurasia. Interestingly, while our strains showed only (Grubisha et al. 2007). At the opposite, spores of ECM one to two diverging bases in ITS sequences (GenBank

Accession nos EU076450–EU076454; M-A. Selosse and L. meiosis) and is thus autogamous (Diez 2005). There are, never- Vincenot, unpublished data), their consensus sequence thless, allogamous species with tetrasporic basidia showed 47 differences, including indels, with a Japanese L. (haploid spores) among host-specific ECM species, such amethystina ITS (AB211270). Therefore, genetic differen- as Laccaria masonii, L. fibrillosa, L. lilacina (specific to Nothofagus tiation in L. amethystina at a continental scale remains spp. in New Zealand) and L. galerinoides (specific to Noth- open to investigation. ofagus spp. in southern America; G. Mueller, personal communication). Thus, although it is worth testing whether autogamy (vs. allogamy) favours evolution of L. amethystina as a multihost species specificity, and whether autogamy is over-represented In hyperdiverse ECM fungal communities (e.g. Kennedy among specific ECM fungi, allogamy is not a strict barrier et al. 2003; Richard et al. 2005; Twieg et al. 2007), special- for evolution of specificity. A second concern is that, due ization, and thus host preferences, can be expected if high to outcrossing, neutral markers could be exchanged, competition favours resource partitioning (MacArthur even under limited gene flow among partially isolated & Pianka 1966). Host specificity and host jumps likely host-specific subpopulations. One might imagine that a low contributed to evolution of several ECM fungal taxa flow homogenizes microsatellite polymorphism (a single (Kretzer et al. 1996; den Bakker et al. 2004), in some of biological species) among populations, whereas, within which morphological differences are less obvious than population, loci involved in symbiosis undergo strong molecular ones (Aanen et al. 2000; Taylor et al. 2007). However, selection by their host, forming host-specific ‘subspecies’. we found no evidence for host-specific CBSs within L. Recently, extensive gene flow in the ECM Suillus luteus was amethystina. Since populations are not strongly isolated shown to impair genetic differenti-ation, based on micro- by distance, any isolation due to host tree could be seen satellites, of populations inhabiting polluted soils (Muller as an evidence for specialization. However, no such trend et al. 2007). This scenario predicts that relevant (selected) was observed (Fig. 2; Table 5), supporting the idea that L. loci in L. amethystina would be polymorphic among amethystina is a true multihost species, over a host range populations under different hosts, but monomorphic in including native trees (Fagus sylvatica, Abies alba, Carpinus populations under a given host, at least if the population betulus and Quercus robur) and the introduced chestnut is sufficiently old for selection to have acted. The related tree (Castanea europaea). Furthermore, assignation tests L. bicolor is a promising model to find relevant loci, or even performed on fruitbodies from the small populations loci portable to L. amethystina, since many genes important (Oa and Ofq) failed to indicate any host-related pattern. for the ECM symbiosis are known in L. bicolor (Kropp & The same result was obtained for assignation of fruitbodies Mueller 1999) and since its genome is fully sequenced from another set of 13 small French populations found (Martin et al. 2008). under various host trees (same hosts and the more recently For ECM fungi, the costs of being generalist (such as introduced Pseudotsuga menziesii and Picea abies, at other underperforming with each host as compared to specific geographical sites; not shown). L. amethystina seemed a species; Kawecki 1994) might be counterbalanced by spa- good candidate for encompassing biological species, since tial and temporal heterogeneity of host communities in in the Laccaria genus, species delineation remains hitherto temperate forests. First, unmanaged temperate forests unclear, especially in Europe where the morphospecies contain multiple species that are heterogeneously distrib- concept mainly applies (Mueller 1992; Kropp & Mueller uted and thus spatially diverse (Brokaw & Busing 2000; 1999). Other Laccaria spp. likely encompass unrecognized McCarthy 2001). Below ground, even stronger heterogene- CBSs (P-A. Moreau and G. Mueller, personal community is expected at the root level, so that at the mycelial ication). In addition, European L. amethystina are late-stage scale, generalist species might colonize roots and survive fungi (i.e. occur in old forests where host trees have better than specific ones. Such a heterogeneous host modified the environment and where evidence suggests environment is very different from that of parasites that that, perhaps in adaptation to these conditions, more live mostly or exclusively inside their hosts; this might host-specific fungi are selected; Last et al. 1987; Horton limit the trend to host specialization among mycorrhizal et al . 2005). fungi. Interestingly, the monospecific stands sampled Nevertheless, the demonstration that L. amethystina is in this study resulted from forest management. Second, largely outcrossing raises two concerns. First, allogamy temporal variations also occur in the composition of tree might limit evolution of specificity, since random mating communities, e.g. after gap formation (McCarthy 2001) or impedes the accumulation, at independent loci, of alle- during ecological succession (e.g. Richard et al. 2005). les selectively useful for interaction with a given host Generalist fungi may thus survive better over long periods (recombination load). Noteworthy, the Eucalyptus-specific in these cases. More generally, extant forest management Laccaria fraterna species has bisporic basidia (each spore practices favouring monospecific, undisturbed stands receives two compatible haploid nuclei resulting from might favour specific ECM species; this could drive some

© 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd 12 M. ROY ET AL . host specialization over a few generations, as recently Belkhir K et al. (1996) GENETIX, Logiciel sous Windows pour la described for parasitic mites forced to live on different hosts Génétique des Populations. Laboratoire Génome, Populations, (Magalhaes et al. 2007). Then, secondary genetic isolation Interactions. CNRS UMR 5000, Université de Montpellier II, Montpellier, France. might be selected to avoid mixing of traits linked to Bergemann SE, Douhan GW, Garbelotto M, Miller SL (2006) No specificity (Kohn 2005). Evolution of ECM specificity in evidence of population structure across three isolated sub- managed forests therefore deserves further attention. populations of Russula brevipes in an oak/pine woodland. New Phytologist, 170, 177–184. Bergemann SE, Miller SL (2002) Size, distribution, and persistence Outlook: networks? of genets in local populations of the late-stage ectomycorrhizal Gene flow in field populations provided evidence that basidiomycete, Russula brevipes. New Phytologist, 156, 313–320. Bever JD (2002) Negative feedback within a mutualism: host- an ECM species can be truly multihost in situ. This result specific growth of mycorrhizal fungi reduces plant benefit. supports the common view that ectomycorrhizal fungi are Proceedings of the Royal Society B: Biological Sciences, 269, 2595–2601. not host-specific and adds to the growing genetic evidence Brokaw N, Busing RT (2000) Niche versus chance and tree diver- that common mycelial networks link different plant sity in forest gaps. Trends in Ecology & Evolution, 15, 183–188. species (Selosse et al. 2006). So far, genetic evidence that a Bruns TD, Bidartondo MI, Taylor DL (2002) Host specificity in given genet colonizes several trees is limited to trees from ectomycorrhizal communities: what do the exceptions tell us? the same species (Lian et al. 2006). Evidence that a given Integrative and Comparative Biology, 42, 352–359. Chevenet F, Brun C, Banuls AL, Jacq B, Christen R (2006) treedyn: ECM genet occurs on several hosts is up to now restricted towards dynamic graphics and annotations for analyses of to achlorophyllous species that exploit common mycelial trees. BMC Bioinformatics, 7. networks to obtain carbon from surrounding plants (Taylor Desmarais E, Lanneluc I, Lagnel J (1998) Direct amplification of & Bruns 1997; Selosse et al. 2002). Although fungal sharing length polymorphisms (DALP), or how to get and characterize between plant species might have important conse- new genetic markers in many species. Nucleic Acids Research, 26, quences, the frequency of this phenomenon still awaits 1458–1465. rigorous analysis. Diez J (2005) Invasion biology of Australian ectomycorrhizal fungi introduced with eucalypt plantations into the Iberian Peninsula. Biological Invasions, 7, 3–15. Acknowledgements Douhan GW, Rizzo DM (2005) Phylogenetic divergence in a local population of the ectomycorrhizal fungus Cenococcum The authors warmly thank Catherine Reeb, Nicolas Mathevon geophilum. New Phytologist, 166, 263–271. and his family, and the Office National des Forêts (division de Dunham SM, O’Dell TE, Molina R (2006) Forest stand age and the Chantilly) for help during fruitbody sampling, as well as Marie- occurrence of chanterelle (Cantharellus) species in Oregon’s cen- Catherine Boisselier-Dubaye, Sarah Samadi and Anne-Laure tral Cascade Mountains. Mycological Research, 110, 1433–1440. Robert for help with microsatellite design at the Muséum National Excoffier L, Schneider S, Roessli D (2000) ARLEQUIN: A Software d’Histoire Naturelle (Service de Systématique Moléculaire). They for Population Genetics Data Analysis, Genetics and Biometry thank Pierre-Arthur Moreau, Greg Mueller, Franck Richard, Laboratory, University of Geneva. Catherine Zabinski and three anonymous reviewers for rich Fiore-Donno A-M, Martin F (2001) Populations of ectomycorrhizal discussions. They thank Simon Tillier and François Le Tacon for Laccaria amethystina and Xerocomus spp. show contrasting their support in launching this work, funded by the Institut colonization patterns in a mixed forest. New Phytologist, 152, Français de la Biodiversité (‘Distribution de la biodiversité chez 533–542. quatre champignons symbiotiques’) and the Centre National de la Geml J, Laursen GA, O’Neill K, Nusbaum HC, Taylor DL (2006) Recherche Scientifique (ATIP to M-A. Selosse). Most molecular Beringian origins and cryptic speciation events in the fly agaric data used in this work were produced through molecular genetic (Amanita muscaria). Molecular Ecology, 15, 225–239. analysis technical facilities of the IFR119 ‘Montpellier Environne- Gherbi H, Delaruelle C, Selosse M-A, Martin F (1999) High diver- ment Biodiversité’. sity in a population of the ectomycorrhizal basidiomycete Laccaria amethystina in a 150-year-old beech forest. Molecular Ecology, 8, 2003–2013. References Gonthier P, Garbelotto M, Varese GC, Nicolotti G (2001) Relative Aanen DK, Kuyper TW, Mes THM, Hoekstra RF (2000) The evo- abundance and potential dispersal range of intersterility groups lution of reproductive isolation in the ectomycorrhizal Hebeloma of Heterobasidion annosum in pure and mixed forests. Canadian crustuliniforme aggregate (basidiomycetes) in Northwestern Journal of Botany, 79, 1057–1065. Europe: a phylogenetic approach. Evolution, 54, 1192–1206. Grubisha LC, Bergemann SE, Bruns TD (2007) Host islands within den Bakker HC, Zuccarello GC, Kuyper TW, Noordeloos ME the California Northern Channel Islands create fine-scale genetic (2004) Evolution and host specificity in the ectomycorrhizal structure in two sympatric species of the symbiotic ectomycor- genus Leccinum. New Phytologist, 163, 201–215. rhizal fungus Rhizopogon. Molecular Ecology, 16, 1811–1822. de la Bastide PY, Kropp BR, Piché Y (1995) Vegetative interactions Grünig CR, Brunner PC, Duo A, Sieber TN (2007) Suitability among mycelia of Laccaria bicolor in pure culture and in sym- of methods for species recognition in the Phialocephala fortinii– biosis with Pinus banksiana. Canadian Journal of Botany, 73, Acephala applanata species complex using DNA analysis. 1768–1779. Fungal Genetics and Biology, 44, 773–788.

Gryta H, Debaud J-C, Marmeisse R (2000) Population dynamics of MacArthur RH, Pianka ER (1966) On optimal use of a patchy the symbiotic mushroom Hebeloma cylindrosporum: mycelial environment. American Naturalist, 100, 603–&. persistence and inbreeding. Heredity, 84, 294–302. Magalhaes S, Fayard J, Janssen A, Carbonell D, Olivieri I (2007) Gryta H, Carriconde F, Charcosset JY, Jargeat P, Gardes M (2006) Adaptation in a spider mite population after long-term evolution Population dynamics of the ectomycorrhizal fungal species on a single host plant. Journal of Evolutionary Biology, 20, 2016–2027. Tricholoma populinum and Tricholoma scalpturatum associated Martin F, Aerts A, Ahrén D et al. (2008) The genome of Laccaria bicolor with black poplar under differing environmental conditions. provides insights into mycorrhizal symbiosis. Nature, 452, 88–92. Environmental Microbiology, 8, 773–786. Martin F, Diez J, Dell B, Delaruelle C (2002) Phylogeography of Horton TR (2002) Molecular approaches to ectomycorrhizal diversity the ectomycorrhizal Pisolithus species as inferred from nuclear studies: variation in ITS at a local scale. Plant and Soil, 244, 29–39. ribosomal DNA ITS sequences. New Phytologist, 153, 345–357. Horton TR, Molina R, Hood K (2005) Douglas-fir ectomycorrhizae McCarthy J (2001) Gap dynamics of forest trees: a review with par- in 40- and 400-year-old stands: mycobiont availability to late ticular attention to boreal forests. Environmental Review, 9, 1–59. successional western hemlock. Mycorrhiza, 15, 393–403. Moore D, Novak Fraser L (2002) Essential Fungal Genetics. Jany JL, Bousquet J, Gagne A, Khasa DP (2006) Simple sequence Springer-Verlag, New York. repeat (SSR) markers in the ectomycorrhizal fungus Laccaria Moreau PA, Peintner U, Gardes M (2006) Phylogeny of the bicolor for environmental monitoring of introduced strains and ectomycorrhizal mushroom genus Alnicola (Basidiomycota, molecular ecology applications. Mycological Research, 110, 51–59. Cortinariaceae) based on rDNA sequences with special Johannesson H, Stenlid J (2004) Nuclear reassortment between emphasis on host specificity and morphological characters. vegetative mycelia in natural populations of the basidiomycete Molecular Phylogenetics and Evolution, 38, 794–807. Heterobasidion annosum. Fungal Genetics and Biology, 41, 563–570. Mueller GM (1992) Systematics of Laccaria (Agaricales) in the Kauserud H, Schumacher T (2003) Genetic structure of fennoscan- Continental United States and Canada, with discussions on dian populations of the threatened wood-decay fungus Fomitopsis extralimital taxa and descriptions of extant types. Fieldiana 30. rosea (Basidiomycota). Mycological Research, 107, 155–163. Field Museum of Natural History, Chicago. Kawecki TJ (1994) Accumulation of deleterious mutations and the Muller LAH, Vangronsveld J, Colpaert JV (2007) Genetic structure evolutionary cost of being a generalist. American Naturalist, 144, of Suillus luteus populations in heavy metal polluted and non- 833–838. polluted habitats. Molecular Ecology, 16, 4728–4737. Kennedy PG, Izzo AD, Bruns TD (2003) There is high potential Murat C, Diez J, Luis P et al. (2004) Polymorphism at the ribosomal for the formation of common mycorrhizal networks between DNA ITS and its relation to postglacial re-colonization routes of the understorey and canopy trees in a mixed evergreen forest. Jour- Perigord truffle Tuber melanosporum. New Phytologist, 164, 401–411. nal of Ecology, 91, 1071–1080. Perrier X, Flori A, Bonnot F (2003) Data analysis methods. In: Kohn LM (2005) Mechanisms of fungal speciation. Annual Review Genetic Diversity of Cultivated Tropical Plants (eds Hamon P, of Phytopathology, 43, 279–308. Seguin M, Perrier X, Glaszmann JC), pp. 43–76. Science Pub- Kretzer A, Li YN, Szaro T, Bruns TD (1996) Internal transcribed lishers, Enfield, New Hampshire. spacer sequences from 38 recognized species of Suillus sensu lato: Rousset F (2008) genepop′007: a complete re-implementation of Phylogenetic and taxonomic implications. Mycologia, 88, 776–785. the genepop software for Windows and Linux. Molecular Ecology Kretzer A, Dunham S, Molina R, Spatafora JW (2005) Patterns of Notes, 8, 103–106. vegetative growth and gene flow in Rhizopogon vinicolor and Richard F, Millot S, Gardes M, Selosse M-A (2005) Diversity and R. vesiculosus (Boletales, Basidiomyceta). Molecular Ecology, 161, specificity of ectomycorrhizal fungi retrieved from an old- 313–323. growth Mediterranean forest dominated by Quercus ilex. New Kropp BR, Mueller GM (1999) Laccaria. In: Ectomycorrhizal Fungi: Phytologist, 166, 1011–1023. Key Genera in Profile (eds Cairney JWG, Chambers SM), Saitou N, Nei M (1987) The neighbor-joining method — a new pp. 65–88. Springer-Verlag, Berlin Heidelberg. method for reconstructing phylogenetic trees. Molecular Biology Last FT, Dighton J, Mason PA (1987) Successions of sheathing and Evolution, 4, 406–425. mycorrhizal fungi. Trends in Ecology & Evolution, 2, 157–161. Sarthou C, Boisselier-Dubayle MC, Lambourdiere J, Samadi S Le Gac M, Hood ME, Fournier E, Giraud T (2007) Phylogenetic (2003) Polymorphic microsatellites for the study of fragmented evidence of host-specific cryptic species in the anther smut populations of Pitcairnia geyskesii L. B. Smith (Bromeliaceae), a fungus. Evolution, 61, 15–26. specific saxicolous species of inselbergs in French Guiana. Lian CL, Narimatsu M, Nara K, Hogetsu T (2006) Tricholoma Molecular Ecology Notes, 3, 221–223. matsutake in a natural Pinus densiflora forest: correspondence Selosse M-A, Costa G, Battista CD, Tacon FL, Martin F (1996) between above- and below-ground genets, association with Meiotic segregation and recombination of the intergenic spacer multiple host trees and alteration of existing ectomycorrhizal of the ribosomal DNA in the ectomycorhizal basidiomycete communities. New Phytologist, 171, 825–836. Laccaria bicolor. Current Genetics, 30, 332–337. Liang Y, Guo LD, Ma KP (2004) Genetic structure of a population Selosse M-A, Martin F, Tacon FL (1998) Survival of an introduced of the ectomycorrhizal fungus Russula vinosa in subtropical ectomycorrhizal Laccaria bicolor strain in European forest plan- woodlands in southwest China. Mycorrhiza, 14, 235–240. tation monitored by mitochondrial ribosomal DNA analysis. Liang Y, Guo LD, Ma KP (2005) Population genetic structure of an New Phytologist, 140, 753–761. ectomycorrhizal fungus Amanita manginiana in a subtropical Selosse MA, Weiss M, Jany JL, Tillier A (2002) Communities and forest over two years. Mycorrhiza, 15, 137–142. populations of sebacinoid basidiomycetes associated with the LoBuglio KF, Taylor JW (2002) Recombination and genetic dif- achlorophyllous orchid Neottia nidus-avis (L.) LCM Rich. and feentiation in the mycorrhizal fungus Cenococcum geophilum neigh-bouring tree ectomycorrhizae. Molecular Ecology, 11, 1831– Fr. Mycologia, 94, 772–780. 1844.

Selosse MA, Richard F, He XH, Simard SW (2006) Mycorrhizal networks: des liaisons dangereuses? Trends in Ecology & Evolution, This work is part of M. Roy’s thesis on mycorrhizal networks that 21, 621–628. link different plant species and their role in plant physiology. M. Smith SE, Read DJ (1997). Mycorrhizal Symbiosis. Academic Press, Roy has been doing the molecular typing and genetic analyses. M. London. Proffit initiated the microsatellite screening during her master of Taylor DL, Bruns TD (1997) Independent, specialized invasions of science. M–P. Dubois, L. Vincenot and E. Desmarais gave help in ectomycorrhizal mutualism by two nonphotosynthetic orchids. the molecular techniques used in this study. This project was Proceedings of the National Academy of Sciences of the United States designed by M-A. Selosse, in a program on ecological role of of America, 94, 4510– 4515. mycorrhizal networks, and he sampled the fungal material. M-A. Taylor AFS, Hills AE, Simonini G, Munoz JA, Eberhardt U (2007) Selosse, L. Vincenot and M. Roy wrote the paper. Xerocomus silwoodensis sp nov., a new species within the European X. subtomentosus complex. Mycological Research, 111, 403–408. Taylor JW, Jacobson DJ, Kroken S et al. (2000) Phylogenetic species recognition and species concepts in fungi. Fungal Genetics and Supplementary material Biology, 31, 21–32. The following supplementary material is available for this Taylor JW, Turner E, Townsend JP, Dettman JR, Jacobson D (2006) Eukaryotic microbes, species recognition and the geographic lim- article: its of species: examples from the kingdom Fungi. Philosophical Transactions of the Royal Society B: Biological Sciences, 361, 1947–1963. Fig. S1a,b Primers designed for this study. Tedersoo L, Pellet P, Koljalg U, Selosse MA (2007) Parallel evolutionary paths to mycoheterotrophy in understorey Ericaceae Fig. S2 Correspondence factorial analysis of the 497 fruitbodies and Orchidaceae: ecological evidence for mixotrophy in based on microsatellites and IGS genotypes. Pyroleae. Oecologia, 151, 206–217. Twieg BD, Durall DM, Simard SW (2007) Ectomycorrhizal fungal Fig. S3 Mantel test on the 11 populations under study. succession in mixed temperate forests. New Phytologist, 176, 437– 447. Wadud MA, Lian CL, Nara K, Ishida TA, Hogetsu T (2006) Devel- This material is available as part of the online article from: opment of microsatellite markers from an ectomycorrhizal fun- http://www.blackwell-synergy.com/doi/abs/ gus, Laccaria amethystina, by a dual-suppression-PCR technique. 10.1111/j.1365-294X.2008.03790.x Molecular Ecology Notes, 6, 130–132. (This link will take you to the article abstract). Xu J, Cheng M, Tan Q, Pan Y (2005) Fungal mitochondrial inheritance. In: Evolutionary Genetics of Fungi (ed. Xu J), pp. 221–252. Please note: Blackwell Publishing are not responsible for the con- Horizon Bioscience, Norfolk, UK. tent or functionality of any supplementary materials supplied by Zhou Z, Miwa M, Hogetsu T (2001) Polymorphism of simple the authors. Any queries (other than missing material) should be sequence repeats reveals gene flow within and between ecto- directed to the corresponding author for the article. mycorrhizal Suillus grevillei populations. New Phytologist, 149, 339–348.

© 2008 The Authors Journal compilation © 2008 Blackwell Publishing Ltd 79.(1*= 79.(1*= = = _= ` =_=`=_= `= _= `=_= `=_= `=_= `=_= :_= _= (942>(477-.?&1= 34(>'* = 85*(.*8= &884(.&9*= <.9-= 9-*= 2>(4-*9*749745-.(= 47(-.)= 5.54,.:2=&5->11:2 =':9=349=<.9-=.98=&8*=:&1=5745&,:1*8_= 33&18=4+= 49&3>`= ,**3`=.3= 57*88_= = &= 51:5&79= )*8= 47(-.)S*8= 2>(4-S9S749745-*8= 8*= 7S5&79.88*39= *3= 7S,.43= 9745.(&1*`= 84:;*39=*3=8.*=):=:)889`=(*5*3)&39`=5*:=)S9:)*8=5479*39=8:7=1*8=2>(4-S9S749745-*8=)*= (*8=7S,.438=*9=348=(433&.88&3(*8=8:7=1&=2>(4-S9S749745-.*=5479*39=8:794:9=8:7=)*8=47(-.)S*8= 9*25S7S*8`= 9*11*= 6:*= *499.&= 3.):88&;.8 = 4:= 47&1147-.?&= 97.+.)& _= -*?= 1*8= *85T(*8= ):= ,*37*= 47&1147-.?& `=94:9*8=8439=97T8=85S(.+.6:*2*39=&884(.S*8=A=)*8=(-&25.,3438=*(942>(47-.?.*38`= (422*= 94:9*8= 1*8= 2>(4-S9S749745-*8= 9*25S7S*8_= 3= 54:77&.9= &1478= 8&99*3)7*= A= )*= +479*8= (4397&.39*8= 5->14,S3S9.6:*8= 8:7= 1.)*39.9S= )*8= 5&79*3&.7*8= 2>(47-.?.*38`= 2&.8= (*99*= ->549-T8*=7*89*=).++.(.1*=(&7=8*:18=1*8=,*37*8= 5.54,.:2 =*9= *5-&1&39-*7& =(425479*39=A=1&=+4.8= )*8=*85T(*8=2>(4-S9S749745-*8=*3=2.1.*:=9*25S7S=*9=9745.(&1_= 5.54,.:2= 748*:2 = &= S9S= 7S(*22*39=S9:).S*= &:= 5&543`= A= 1&= 1.2.9*= 347)= )*= 843= &.7= )*= 7S5&79.9.43= *9= (*99*= 47(-.)S*= 2>(4-S9S759745-*8= *89= 85S(.+.6:*2*39= &884(.S*= A= )*8= 457.3&(*&*`= )*8= (-&25.,3438= 8&5745->9*8_= 6&:97*8= 47(-.)S*8= 9745.(&1*8= 8439= &884(.S*8= 85S(.+.6:*2*39= A= )*8= (-&25.,3438= 8&5745->9*8`= (422*= &1*41&= &19.88.2& = 4: = &8974).&= 8.0040.&3& _= *99*= &884(.&9.43= A= )*8= 8&5745->9*8= *3= 2.1.*:= 9745.(&1= *898*11*= 57457*= A= (*= 2.1.*:=4:=A=(*8=*85T(*8=*3=5&79.(:1.*7== _=748*:2 =*89=1*85T(*=8":7=) _=&5->11:2 `=:3*=*85T(*= *:7&8.&9.6:*=51:99=9*25S7S*=6:4.6:*=7&7*2*39=;:*=*3=2.1.*:=9745.(&1_= 3=5*:9=8*=)*2&3)*7= 8.=(*99*=*85T(*=*89=&884(.S*=A=)*8=(-&25.,3438=8.2.1&.7*8=A= _=748*:2 `=&+.3=)S9:).*7=8.=)&38= )*8= *3;.7433*2*398= (4251T9*2*39= ).++S7*398= ):= 54.39= )*= ;:*= )*8= (422:3&:9S8= )*= (-&25.,3438`=(*8=*85T(*8=8439=&884(.S*8=&:==2U2*8=(-&25.,3438_=.=(*=3*89=5&8=1*=(&8`=43= 5*:9= (425&7*7= 1*8= (-&25.,3438= )*= _= 748*:2 = *9= _= &5->11:2 = &:= 5&543`= 5:.86:*= 1*8= )*:== >= *=.89*39= &+.3= )*= 8&;4.7= 8.= 1*8= ).++S7*3(*8= 8439= 51:99= ):*8= A= :3= (-&3,*2*39= )*= 2.1.*:= 4:= 8*:1*2*39= A= )*= 1&= ).++S7*3(.&9.43= 5&7= 1&= ).89&3(*_= 6*= 51:8`= (*8= (425&7&.8438= 5*72*997439= )S9:).*7= 1*8= ;&7.&9.438= ):= (479T,*= )*= (-&25.,3438= 8:7= )*= 97T8= ,7&3)*8= ).89&3(*8= ,S4,7&5-.6:*8_= *8= (-&25.,3438= )*= _= &5->11:2 = 8*= 8439= 7S;S1S8= U97*= 94:8= )*8= (-&25.,3438= *(942>(47-.?.*38=):=,*37*= 34(>'* `=&55&79*3&39=A=51:8.*:78=*85T(*8`=57.3(.5&1*2*39= 34(>'*= +:8(.):1& _=*8=(-&25.,3438=3*=57S8*39&.*39=5&8=)*=7*,74:5*2*39=,S4,7&5-.6:*=*3=7*1&9.43= &;*(=1*:7=548.9.43=5->14,S3S9.6:*=*9=1*8=5&79*3&.7*8=)*8= _=&5->11:2= /&543&.8=S9&.*39=574(-*8= )*=(*:==)*8= _=&5->11:2 =*:745S*38`=*9=343=)*=(*:==)*8= _=748*:2 =/&543&.8_= *8= 7S8:19&98= (43+.72*39= )43(= 1&= 9*3)&3(*= )*8= 2>(4-S9S749745-*8= 9*25S7S*8= A= 8&884(.*7= 97T8= 85S(.+.6:*2*39= A= )*8= (-&25.,3438= *(942>(47-.?.*38= *9= )S24397*= 6:*= )*8= (-&3,*2*398=)-9*=7&).(&:==5*:;*39=&;4.7=1.*:`=94:9=(422*=1478=):=5&88&,*=)*=1S9&9=;*79=A= 1S9&9=2>(4-S9S749745-*_=6&:97*=5&79`=(*8=7S8:19&98=8439=.39S7*88&398=54:7=1&=,*89.43=)*8=7&7*8= 545:1&9.438=)*= _=&5->11:2`= )439=1&='.414,.*=S9&.9=5*:=(433:*_=*9=&79.(1*=7S;T1*=1*=1.*3=3:97.9.+= &:== &7'7*8=6:.= 1*394:7*`= *9= )S24397*=6:*= (*99*=47(-.)S*= 8:7;.9= 57.3(.5&1*2*39= ,7B(*= A= 8*8= (&5&(.9S8=)*=2:19.51.(&9.43=;S,S9&9.;*_= Annals of Botany Page 1 of 16 doi:10.1093/aob/mcn269, available online at www.aob.oxfordjournals.org

Ectomycorrhizal Inocybe species associate with the mycoheterotrophic orchid Epipogium aphyllum but not its asexual propagules

Melanie Roy 1,†, Takahiro Yagame 2, Masahide Yamato 3, Koji Iwase 4, Christine Heinz 5, Antonella Faccio 6, Paola Bonfante 6 and Marc-Andre Selosse 1, †,* 1Centre d’Ecologie Fonctionnelle et Evolutive (CNRS, UMR 5175), Equipe Interactions Biotiques, 1919 Route de Mende, 34293 Montpellier ce´dex 5, France, 2Orchid Museum Takamori, 512-73 Izuhara, Shimoina, Nagano 399-3107, Japan, 3Environment Department, The General Environmental Technos Co., Ltd, 1-3-5 Azuchimachi Chuo-ku, Osaka 541-0052, Japan, 4Fungus/ Mushroom Resource and Research Center, Faculty of Agriculture, Tottori University, 4-101 Koyama-minami, Tottori 680-8553, Japan, 5Universite´ Montpellier 2, UMR AMAP Botanique et bioinformatique de l’Architecture des Plantes, 34000 Montpellier, France and 6Dipartimento di Biologia Vegetale dell’Universita`, Istituto per la Protezione delle Piante – CNR, Viale Mattioli 25, 10125 Torino, Italy

Received: 1 June 2008 Returned for revision: 22 September 2008 Accepted: 25 November 2008

† Background and Aims Epipogium aphyllum is a Eurasian achlorophyllous, mycoheterotrophic forest orchid. Due to its rarity, it is often protected, and its biology is poorly known. The identity and pattern of colonization of fungal associates providing carbon to this orchid have not been studied previously. † Methods Using samples from 34 individuals from 18 populations in Japan, Russia and France, the following were investigated: ( a) colonization patterns of fungal associates of E. aphyllum by microscopy; ( b) their identity by PCR amplification of nuclear ribosomal ITS carried out on rhizome fragments and hyphal pelotons. † Results and Conclusions Microscopic investigations revealed that thick rhizomes were densely colonized by fungi bearing clamp-connections and dolipores, i.e. basidiomycetes. Molecular analysis identified Inocybe species as exclusive symbionts of 75 % of the plants investigated and, more rarely, other basidiomycetes (Hebeloma , Xerocomus , Lactarius , Thelephora species). Additionally, ascomycetes, probably endophytes or parasites, were sometimes present. Although E. aphyllum associates with diverse species from Inocybe subgenera Mallocybe and Inocybe sensu stricto , no evidence for cryptic speciation in E. aphyllum was found. Since basidiomycetes colonizing the orchid are ectomycorrhizal, surrounding trees are probably the ultimate carbon source. Accordingly, in one population, ectomycorrhizae sampled around an individual orchid revealed the same fungus on 11 .2 % of tree roots investigated. Conversely, long, thin stolons bearing bulbils indicated active asexual multiplication, but these propagules were not colonized by fungi. These findings are discussed in the framework of ecology and evolution of mycoheterotrophy.

Key words: Asexual multiplication, ectomycorrhizae, Epipogium , Inocybe , mycoheterotrophy, orchid mycorrhizae, speciﬁcity, symbiont transmission.

INTRODUCTION orchids has used molecular techniques to identify fungal symbionts that are often unculturable and thus unidentifiable from Orchids depend on their fungal symbionts at germination since their morphology in vitro (Taylor et al. , 2002; Dearnaley, their seeds are devoid of food reserves. Soil fungi that colonize 2007). MH fungal symbionts turned out to differ from the orchid seeds provide carbon and mineral resources and allow usual rhizoctonias, both at taxonomic and ecological levels. their development into a heterotrophic, underground proto- Most MH orchids associate with basidiomycetes that also corm. These fungal symbionts usually belong to a few unre- form so-called ectomycorrhizae (ECM) on roots of trees and lated basidiomycete taxa collectively called ‘rhizoctonias’ shrubs (Smith and Read, 1997). Russulaceae were found in (mostly Ceratobasidiaceae, Tulasnellaceae and some Corallorhiza maculata (Taylor and Bruns, 1997, 1999), Sebacinales; Rasmussen, 2002). Adult orchids are often auto- C. mertensiana (Taylor and Bruns, 1999), Limodorum trophic and still harbour fungi in their roots, forming typical species (Girlanda et al. , 2006), Dipodium variegatum mycorrhizal associations (Smith and Read, 1997) and (Bougoure and Dearnaley, 2005) and D. hamiltonianum perhaps reversing the carbon flow toward fungi (Cameron (Dearnaley and Le Brocque, 2006). Thelephoraceae occurred et al. , 2006). However, during orchid evolution, photosyn- in Cephalanthera austinae (Taylor and Bruns, 1997), thesis was lost .20 times (Molvray et al. , 2000), and non- Corallorhiza trifida (Taylor and Bruns, 1997; McKendrick green, ‘mycoheterotrophic’ (MH) orchids receiving carbon et al. , 2000) and C. striata (Taylor et al. , 2002). Sebacinales from their fungal symbionts at adult stage have evolved were found in Neottia nidus-avis (McKendrick et al. , 2002; convergently (Leake, 1994, 2004). Recent research on MH Selosse et al. , 2002 b) and Hexalectris spicata (Taylor et al. , * For correspondence. E-mail [email protected] 2003); although rhizoctonias encompass some Sebacinales, †These two authors equally contributed to this work. the species from MH orchids belonged to a different clade # The Author 2009. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For Permissions, please email: [email protected] Page 2 of 16 Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species with ECM abilities (clade A in Weiss et al. , 2004). Using i 14 C-labelling, McKendrick et al. (2000) demonstrated that the shared fungus provided MH orchids with photosynthates from host trees of the fungi. Some rhizoctonias belonging to Ceratobasidiaceae also associate with the MH Rhizanthella gardneri (Warcup, 1991) and Chamaegastrodia sikokiana (Yagame et al. , 2008); in both cases, the isolated fungi were able to form ECM in vitro , and these orchids probably also depend on photosynthates from the trees. However, there are also some reports of saprobic basidiomy- 5 mm cetes in MH orchids, such as Armillaria species in Galeola septentrionalis (Cha and Igarashi, 1996) and Gastrodia elata (Lan et al. , 1994) or Erythromyces species in Galeola altis- is sima (Umata et al. , 1995) and Erythrorchis cassythoides 1 cm (Dearnaley, 2006). In fact, Armillaria mellea colonizes and is induces growth of seedlings of Galeola septentrionalis (Terashita, 1985), and Erythromyces crocicreas induces germi- ib ib nation of Galeola altissima (Umata, 1995). Campbell (1970) sb isolated several saprobic fungi from MH orchids. However, s s these saprobic fungi were identified after in vitro isolation, and saprobic contaminants may have overgrown the true s mycorrhizal fungi that are often slow growing or unculturable (Taylor et al. , 2002). Indeed, fungi isolated by Campbell cr (1970) were not confirmed by recent molecular studies. Molecular methods are also sensitive to contamination by F I G . 1. Morphology of E. aphyllum (redrawn from Irmisch, 1853) and rhizoid-bearing protocorm (inset). cr, Coralloid rhizome; i, inflorescence; ib, DNA from endophytic fungi or spores, so that direct obser- inflorescence bud; is, inflorescence shoot; s, thin stolon; sb, bulbil on a stolon. vations of fungi (e.g. by electron microscopy; Selosse et al. , 2004) or functional tests (such as seed germination using the of hypogeous plant parts has an impact on fungal colonization isolated fungus) are necessary to corroborate the identity of strategies and ( d) how transmission to asexual offspring mycorrhizal fungi. occurs. These data are of relevance to future in situ or ex Recently, fungi belonging to Coprinaceae, a group of sapro- situ conservation activities, especially because this species is bic basidiomycetes, were found in the Asian MH Epipogium rare over its range. For example, it has been considered roseum (Yamato et al. , 2005) and Eulophia zollingeri extinct in the United Kingdom since 1987 (Harrap and (Ogura-Tsujita and Yukawa, 2008). Appropriately, E. roseum Harrap, 2005; Kull and Hitchings, 2006) and is highly pro- grows near tree stumps and fallen logs, and the fungus isolated tected in France (Danton and Baffray, 2005); in Japan, it is from E. roseum allowed in vitro seed germination and develop- listed as an endangered species in the Red Data Book ment up to flowering stage (Yagame et al. , 2007), therefore (Environment Agency of Japan, 2000). fulfilling one of the criteria defining a symbiotic fungus. Epipogium occurs throughout Eurasia and Africa (Pridgeon et al. , 2005) and may thus encompass overlooked associations MATERIALS AND METHODS with saprobic fungi. Alternatively, the other Epipogium Rhizome and surrounding ECM sampling species associate with ECM fungi, in which case Epipogium would illustrate a hitherto unknown variability in ecology of Rhizomes were sampled in August 2005 in three countries fungal partners. where official authorizations were obtained (France, Russia This study focuses on E. aphyllum , which occurs from and Japan), using a protocol that allows plant survival (data Europe to Asia (Maekawa, 1971; Rasmussen, 1995). not shown). One to ten independent rhizome fragments were Questions raised by previous studies on E. roseum are harvested by digging about 20 cm away from shoots and addressed: ( a) What are the taxonomic position and ecology then carefully approaching underground parts of the plant of E. aphyllum symbionts? ( b) How much variation is there from one side; after sampling, the hole was refilled with the in symbionts over the range of the species? Epipogium aphyl- same soil. Up to three coralloid rhizome fragments (‘cr’ on lum has a complex vegetative morphology, forming two kinds Fig. 1A) and, when available, thin stolons (‘s’ on Fig. 1A) of rhizomes (Fig. 1A): thick, highly branched rhizomes (the were collected from one to three plants per population so-called ‘coralloid rhizomes’) and thin stolons, up to 0 .5 m (Table 1). In a large French population at Saint Cle´ment long (Irmisch, 1853; Ziegenspeck, 1936). The latter presum- (Cantal), with .300 shoots, two full plants (EM12 and ably contribute to asexual reproduction. Although root- or EM15; Table 1) situated at 150 m from each other were recov- rhizome-sprouting is common among MH plants (Leake, ered, and all ECM tips of surrounding trees found ,15 cm 1994) and sometimes includes transmission of the fungal sym- from each of these plants were harvested (in all, 71 þ 46 biont (Domıńguez et al. , 2006), the presence of fungi in pro- ECM tips, respectively). In the same population, rhizome frag- pagules has rarely been assessed (Klimesˇova´ et al. , 2007). ments were sampled from seven plants to address intrapopula- Therefore it is also questioned here ( c) whether the anatomy tion diversity. Samples were carefully washed with water to Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species Page 3 of 16

TABLE 1. Identiﬁcation of fungal sequences retrieved from the investigated E. aphyllum individuals, classiﬁed by origin

Closest sequence(s) Genbank found in GenBank Latitude/ accession Gene Possible Fungal by BLAST: name E Maximum Region longitude Individual* number sequenced † afﬁliation ‡ ecology § (accession no.) value } identity (%)

Lans en Vercors, 45 807 047 00 N EM5 EU711167 28S c 2 Olpidium P Olpidium brassicae 8e – 112 95 Alpes 05 835 020 00 E (DQ273818) EU711168 ITS c 4 Olpidium P Olpidium 4e – 71 100 bornovanus (AB205215) EU711169 ITS c 2 Thelephora ECM Uncultured 0 99 Thelephoraceae (AY634145) EU711170 ITS c 2 Inocybe ECM Inocybe cf. 0 98 geophylla (AM882984) EM6 EU711171 ITS þ 28S d 1 Inocybe ECM Inocybe pudica 0 87 (AY228341) EU711172 ITS d 1 Inocybe ECM Inocybe cf. 0 100 geophylla (AM882984) Villard de Lans, 45 804 011 00 N EM7 EU711173 ITS d 1 Inocybe ECM Inocybe terrigena 0 99 Alpes 05 833 00 01 00 E (AM882864) Luz Saint Sauveur, 42 851 054 00 N EM1 EU711163 ITS d 1 Inocybe ECM Inocybe fuscidula 0 96 Pyreńeés 0801 013 00 W (AM882888) Thue`s-entre-Valls, 42 830 057 00 N EM8 EU711174 28S c 2 Inocybe ECM Inocybe 0 98 Pyreńeés Orientales 2815 0900 E griseolilacina (AY380378) EU711175 ITS c 1 Inocybe ECM Inocybe cf. 0 90 glabripes (AJ889952) 42 8 30 030 00 N EM9 EU711177 ITS þ 28S c 1 Protoventuria P Protoventuria alpina 0 96 2815 027 00 E (EU035444) EU711176 ITS þ 28S c 1 Inocybe ECM Inocybe nitidiuscula 0 85 (AM882912) 42 830 018 00 N EM23 EU711209 ITS þ 28S d 1 Inocybe ECM Inocybe fuscidula 0 98 2815 029 00 E (AM882888) EU711210 ITS d 1 Inocybe ECM Inocybe fuscidula 0 97 (AM882888) EM24 EU711211 ITS þ 28S d 2 Inocybe ECM Inocybe fuscidula 0 100 (AM882888) Cazaux, Pyreńeés 43 803 008 00 N EM17 EU711189 ITS þ 28S d 4 Inocybe ECM Inocybe fuscidula 0 98 01 830 035 00 E (AM882888) EU711190 ITS d 1 Inocybe ECM Inocybe fuscidula 4e – 94 89 (AM882888) EU711187 ITS d 1 Inocybe ECM Inocybe fuscidula 0 98 (AM882888) EU711188 ITS d 1 Inocybe ECM Inocybe fuscidula 0 96 (AM882888) EM18 EU722336 ITS þ 28S d 3 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) Ooˆ, Pyreńeés 42 847 049 0N EM20 EU711191 ITS þ 28S d 1 Inocybe ECM Inocybe flocculosa 0 89 0830 021 00 E (AY228354 ) EU711199 ITS d 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711200 ITS d 1 Inocybe ECM Inocybe rufuloides 0 91 (DQ067579) EU711201 ITS d 1 Inocybe ECM Inocybe rufuloides 0 94 (DQ067579) EU711192 ITS d 1 Inocybe ECM Inocybe rufuloides 0 91 (DQ067579) EU711193 ITS d 1 Inocybe ECM Inocybe rufuloides 2e – 165 89 (DQ067579) EU711194 ITS d 1 Inocybe ECM Inocybe nitidiuscula 2e – 124 88 (AM882912) EU711195 ITS d 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) Continued Page 4 of 16 Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species

TABLE 1. Continued

EU711196 ITS d 1 Inocybe ECM Inocybe fuscidula 0 100 (AM882888) EU711197 ITS d 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711198 ITS þ 28S Inocybe ECM Inocybe fuscidula 0 99 d1 þ c2 (AM882888) Saint Cle´ment, 44 855 028 00 N EM2 EU711164 ITS d 1 Inocybe ECM Inocybe dulcamara 0 96 Cantal 02 839 041 00 E (AM882863) EM10 EU711178 28S c 1 Neonectria E Neonectria 0 98 radicicola (AY283552) EU711179 28S c 1 Protoventuria P Protoventuria alpina 0 98 (EU035444) EM11 EU711180 ITS c 1 Metarhizium SS Metarhizium 0 90 anisopliae (AB027383) EU711181 ITS c 1 Paecilomyces SS Paecilomyces 0 100 carneus (AB258369) EM12 EU711182 ITS d 1 Inocybe ECM Inocybe dulcamara 0 99 (AM882863) EM15 EU711184 ITS d 1 Inocybe ECM Inocybe lanuginosa 1e – 138 84 (EU525948) EM16 EU711185 ITS c 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711186 ITS þ 28S c 1 Didymella P Didymella bryoniae 0 84 (AB266850) Saint Paul de Salers, 45 808 025 00 N EM22 EU711202 ITS þ 28S c 1 Inocybe ECM Inocybe whitei 0 86 Cantal 02 831 004 00 E (EU486441) EU711203 ITS d 1 Inocybe ECM Inocybe fuscidula 0 100 (AM882888) EU711204 ITS d 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711205 ITS d 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711206 ITS d 1 Inocybe ECM Inocybe fuscidula 6e – 151 87 (AM882888) EU711207 ITS d 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711208 ITS d 1 Inocybe ECM Inocybe fuscidula 0 100 (AM882888) Picherande, Puy de 45 827 050 00 N EM26 EU711217 ITS þ 28S d 1 Inocybe ECM Inocybe whitei 0 88 Doˆme 02 846 009 00 E (EU486441) EU711216 ITS d 1 Inocybe ECM Inocybe rufuloides 0 92 (DQ067579) EU711213 ITS d 1 þ c1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711214 ITS d 1 þ c1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711215 ITS d 1 Inocybe ECM Inocybe fuscidula 0 100 (AM882888) EU711218 ITS c 1 Neonectria E Neonectria 0 99 radicicola (AJ875331) EM27 EU711219 ITS þ 28S d 1 Inocybe ECM Inocybe armeniaca 0 88 (DQ974803) EU711220 ITS d 1 Inocybe ECM Inocybe rufuloides 0 90 (DQ067579) EU711221 ITS c 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EM28 EU711222 ITS d 1 Inocybe ECM Inocybe fuscidula 7e – 124 92 (AM882888) Continued Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species Page 5 of 16

TABLE 1. Continued

EU711223 ITS d 1 Inocybe ECM Inocybe fuscidula 0 98 (AM882888) EU711224 ITS d 1 Inocybe ECM Inocybe fuscidula 0 96 (AM882888) EU711225 ITS d 1 Inocybe ECM Inocybe fuscidula 0 97 (AM882888) EU711226 ITS d 1 Inocybe ECM Inocybe fuscidula 0 98 (AM882888) Ponteils, Ce´vennes 44 853 031 00 N EM3 EU711165 ITS d 1 Lactarius ECM Lactarius 0 93 03 857 032 00 E scrobiculatus (EF530942) EM4 EU711166 ITS d 1 Inocybe ECM Inocybe geophylla 0 96 (AM882870) RUSSIA Arshan, Byryatia 51 854 022 00 N EM14 EU711183 ITS d 1 Inocybe ECM Inocybe lanuginosa 1e – 169 85 Ural 102 826 002 00 E (EU525948) Murzinka, Lac 51 837 051 00 N EM29 EU711227 28S d 4 Inocybe ECM Inocybe dulcamara 0 99 Baı¨kal 82 832 040 00 E (AY038315) JAPAN Nagano prefecture 35 858 020 00 N NA1 EU711233 ITS p 1 Inocybe ECM Inocybe fuscidula 0 98 138 818 026 00 E (AM882888) EU711234 ITS p 2 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) NA2 EU711235 ITS p 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) 35 858 026 00 N NB1 EU711236 ITS p 1 Inocybe ECM Inocybe fuscidula 0 99 138 818 040 00 E (AM882888) EU711237 ITS p 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) EU711238 ITS p 1 Inocybe ECM Inocybe fuscidula 0 99 (AM882888) 35 858 020 00 N NC1 EU711238 ITS p 1 Inocybe ECM Inocybe fuscidula 0 99 138 820 00 900 (AM882888) Yamanashi 35 824 001 00 N YA1 EU711239 ITS p 1 Hebeloma ECM Hebeloma velutipes 0 99 prefecture 138 841 014 00 E (AF430254) YA2 EU711240 ITS p 1 Hebeloma ECM Hebeloma velutipes 0 99 (AF430254) YA3 EU711241 ITS p 1 Inocybe ECM Inocybe subnudipes 0 97 (AM882809) EU711242 ITS p 1 Inocybe ECM Inocybe subnudipes 0 97 (AM882809) EU711243 ITS p 1 Inocybe ECM Inocybe subnudipes 0 97 (AM882809)

*The individuals used in the ampliﬁcation of plant ITS are underlined. †Abbreviations: c, from a clone; d, from direct sequencing of a rhizome; p, from direct sequencing of a pool of 20 pelotons. Superscript numbers indicate the number of times a clone or a direct sequence was obtained. ‡Only the closest taxonomically informative accession is reported. §Putative ecology: ECM, ectomycorrhizal; P, parasite; E, plant endophyte; SS, soil saprobe. }The BLAST expected value represents the number of sequence matches expected by random chance (the smaller the value, the better the match between our sample sequences and those in the NCBI database).

eliminate all soil particles. All samples were stored in ethanol/ as in Warcup and Talbot (1967), with modiﬁcations as water (3/2, v/v) for transport, and in some populations follows. Surfaces of rhizomes were sterilized by immersion (Table 1) sub-samples of coralloid rhizome fragments were in 70 % ethanol for 30 s and sodium hypochlorite solution con- preserved for microscopic investigations by quick ﬁxation in taining 1 % available chlorine for 30 s. The rhizome was 2.5 % (v/v) glutaraldehyde in a 0 .1 M cacodylate buffer (pH rinsed and cut with a sterilized scalpel into three pieces, 7.2) for 2 h at room temperature and then overnight at 4 8C. each placed in 5 mL of sterilized distilled water in a Petri For the seven plants from Japan, the sampling was performed dish (9 cm in diameter) and crushed with a sterilized glass by collecting one large rhizome fragment in each population rod to disperse the intracellular hyphal pelotons. Twenty and isolating fungal pelotons ( ¼ intracellular hyphal coils) fungal pelotons per rhizome piece were harvested and pooled. Page 6 of 16 Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species

Molecular investigations Fungal identification and phylogenetic analyses DNA extraction and PCR amplification of fungal internal A search for similar sequences was conducted with Blast in transcribed spacers of nuclear ribosomal DNA (ITS) were per- GenBank (National Center for Biotechnology Information, formed as in Selosse et al. (2002 b) using primers ITS1F and http://www.ncbi.nlm.nih.gov). To confirm the phylogenetic ITS4 on coralloid rhizome fragments and ECM tips. ITS frag- position of the fungal symbionts related to the genus ments amplified from coralloid rhizomes were directly Inocybe , the 5’ part of the 28S rDNA was sequenced using sequenced as in Selosse et al. (2002b), except for the primers ITS1F and TW13 as in Selosse et al. (2002 b). The Japanese samples that were handled as in Yamato and Iwase sequences were aligned with selected Inocybe sequences (2008). Whenever direct sequencing was not possible, PCR pro- from GenBank; Cortinarius odoratus (DQ663360), Galerina ducts were cloned as in Selosse et al. (2004), and at least six autumnalis (AY281020), Hebeloma velutipes (AY818351), clones per plant were sequenced. For ECM tips, length poly- H. leucosarx (AB211268), H. pusillum (AB211274) and morphism of ITS was investigated in comparison to the H. mesophaeum (AB327182) were used as outgroups. fungal ITS amplified from the nearby orchid, before and after Similarly, to investigate plant relationships, sequences of enzymatic digestion (RFLP using EcoR I þ Sac I and Hin dIII). Gastrodia elata (EF090607), Nervilia shiriensis (AF521066), RFLP were carried out as in Selosse et al. (2002 a), and only Cranichis revoluta (AF391786) and Orchis militaris ITS fragments identical in length and RFLP patterns to those (AY699977) were downloaded from Genbank as outgroups. from the nearby orchid were sequenced. Further, to ensure After alignment using Bioedit v7.3.0 (Hall, 1999) and absence of usual rhizoctonia orchid symbionts with highly ClustalW (Thompson, 1994), the result was checked by eye derived rDNA sequences, i.e. tulasnelloid and sebacinoid basi- and corrected manually with Bioedit. A bootstrapped diomycetes, additional PCR amplifications were carried out neighbor-joining analysis (Saitou and Nei, 1987) was per- on coralloid rhizome fragments using specific primers formed with PAUP 4.0 (Phylogenetic Analysis Using (ITS4tul for Tulasnellaceae and ITS3S for Sebacinales, as in Parsimony, version 4.0; Swofford, 2004). Genetic distances Selosse et al. , 2004) and positive controls. Whenever ITS were estimated by maximum likelihood using a general time- typing failed at the PCR step, sequence was tentatively obtained reversible model (Lanave et al. , 1984; Rodriquez et al. , 1990), by amplifying the 28S rDNA using the primers Lr0r and Lr5 involving unequal base frequencies and six types of substi- (Vilgalys and Hester, 1990). To ensure relatedness of the tution. This model of DNA substitution was chosen using a E. aphyllum plants used relative to E. roseum , plant nuclear series of hierarchical likelihood-ratio test in Modeltest 3.7 ribosomal ITS sequences were amplified from four populations (Posada and Crandall, 1998). Base frequencies were estimated (Table 1) using the plant-specific primer ITS1P as in Selosse before running the analysis, and 1000 bootstrap replicates were et al. (2002 a). An E. roseum ITS sequence was obtained from performed. Percentages of sequence identity between fungal a plant collected in Java for the Museum National d’Histoire ITS sequences were measured with Bioedit and compared Naturelle of Paris (voucher: Frank, C.W. 729). Sequencing with geographical distances and elevations by a Mantel test was carried out on an ABI PRISM 3130 XL Genetic analyser (using XLstat; Addinsoft, Paris, France) to test whether these (Applied Biosystems, Foster City, CA, USA), using the PCR two factors were correlated to any differences in fungal primers, and sequences from both strands were edited using associates. Sequencher TM 4.6 for MacOS X from Genes Codes (Ann Arbor, MI, USA). Edited sequences (or consensus sequences RESULTS for similar clones) were deposited in GenBank. Conspecificity of the sampled orchids The four populations selected over the range produced ITS Microscope investigations sequences (EU711228 to EU711231) that diverged among After rinsing with fixing buffer (see above), four subpopulations by a maximum of 8 .5 %. The four resulting samples of coralloid rhizome fragments were dehydrated in sequences clustered together (100 % bootstrap) as a sister an ascending series of ethanol solutions to 100 %, incubated group to E. roseum (EU711232) with 80 % support (Fig. 2), in two changes of absolute acetone and infiltrated with supporting the coherence of the investigated taxa in France, Epon-Araldite resin (Hoch, 1986). The resin was polymerized Russia and Japan. for 24 h at 60 8C. Embedded samples were processed for ultra- microtomy: semi-thin sections (0 .5 mm) were stained with 1 % Fungal colonization toluidine blue and ultra-thin (70 nm) sections were counter- stained with uranyl acetate and lead citrate (Reynolds, 1963). The below-ground portion of E. aphyllum is composed of These were used for TEM analyses under a Philips CM10 plagiotropic coralloid rhizomes (Fig. 3A) that at some point transmission electron microscope. Stolon samples were become either an ascending inflorescence or a thin stolon embedded in paraffin before sectioning. Manual transverse (Fig. 3B). Inflorescence buds were filled with starch (not sections (10 mm) were cut with a microtome, differentially shown). These coralloid rhizomes proved to be densely colo- stained with a mixture of safranin O and fast green FCF nized by fungi, with the exception of the meristematic zone (Bryan, 1955), rinsed with distilled water and observed with at their apex, characterized by a whitish colour (Fig. 3A). a light microscope. To look for fungal colonization of Transverse sections showed that the outer cell layers were stolons, thin sections and gently crushed stolons were stained usually not colonized (Fig. 3C). Some isolated hyphae with Trypan Blue (Koske and Gemma, 1989). running from soil to the more internal cortical cells were Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species Page 7 of 16

EU711230 Epipogium aphyllum France 82

EU711231 Epipogium aphyllum France

100 EU711229 Epipogium aphyllum Japan

80 EU711227 Epipogium aphyllum Russia

84 EU711232 Epipogium roseum

100 Gastrodia elata

Nervilia shinensis

Orchis militaris

Cranichis revoluta 0·8

F I G . 2. Clustering of E. aphyllum and E. roseum individuals with respect to some other genera of Epidendroideae; Cranichis revoluta and Orchis militaris (Orchidoideae) are outgroups (ML on an alignment of the ITS; GTR model, 1000 bootstrap; only bootstraps .80 % are shown). occasionally observed (not shown), but their route of pen- shown). In bulbils, cell nuclei were central, and smaller cells etration could not be followed. The vascular bundle and the formed an apical meristem (Fig. 4F, G). No fungal coloniza- air-filled intercellular spaces (Fig. 3C, D) were not infected. tion was seen in bulbils and stolons after staining with The inner cortical cells were filled with pelotons that were Trypan Blue ( n ¼ 14 stolons and n ¼ 33 bulbils, from n ¼ 7 often elongated in one direction; these push the enlarged cell plants; not shown). nucleus towards the periphery (Fig. 3C, D). Direction of elongation varied from one cell to another, allowing some linear hyphae to be longitudinally cut (Fig. 3D and G). TEM Identification of E. aphyllum mycorrhizal fungi investigations confirmed that hyphae occurred in living cells From a total of 34 plants in 18 populations (Table 1), 146 with intact organelles but without starch (not shown). coralloid rhizome fragments and 21 peloton pools were inves- Hyphae were surrounded by the host plasma membrane tigated. Primers specific for Tulasnellaceae and Sebacinales (arrowed in Fig. 3E, G). They consistently showed dolipores ITS never produced any amplicon, whereas the general between cells (Fig. 3E, F) with surrounding perforate reticu- primer pair (ITS1F and ITS4) and/or 28S rDNA primers suc- lum cisternae (the so-called parenthesome; Fig. 3F); clamp cessfully amplified DNA from 128 plants. Direct sequences connections were also seen (Fig. 3G). On the bases of were obtained for rhizome fragments from 21 plants and 13 these cytological features, the fungus is confirmed as a peloton pools from Japanese plants. Among these 79 basidiomycete. sequences, 65 (82 %, from 27 plants) were related to Inocybe Stolons reached up to 50 cm and produced axillary bulbils sequences in GenBank (Table 1). Two plants from one every 2–3 cm (Figs 3B and 4A, B) more or less deeply pro- Japanese population exhibited a fungus related to Hebeloma tected by a sheathing, scaly leaf. This structure is loose and (Fig. 5), whereas Xerocomus and Lactarius species were fragile, and separation from the mother plant occurred easily found once each in a plant from two French populations. No with disturbance. The oldest bulbils were covered with rhi- fungal fragment was amplified from bulbils or stolons ( n ¼ zoids (Fig. 4C). Sections illustrated the contrasting features 24 and n ¼ 12, respectively, from n ¼ 12 plants). of cortical cells of stolons (elongated, often empty, separated To investigate fungal diversity in rhizome fragments for by air-filled intercellular spaces) and those of bulbils which ITS was not directly sequenced, 12 ITS amplification (densely filled with starch and stacked together, Fig. 4D–F). products were cloned. Ten of these revealed one to four Xylem was poorly differentiated in vascular bundles (not Inocybe -related sequences, sometimes associated with Page 8 of 16 Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species

A B

C D

E F G

F I G . 3. Morphology and fungal colonization of coralloid rhizomes: (A) fragment of a coralloid dichotomously branching rhizome; (B) close-up of the underground parts with bases of inflorescence shoots (is), the coralloid rhizome (cr) sometimes giving rise to stolon(s) bearing bulbils (sb); (C) section at low magnification with pelotons (p) in the internal cell layers, and absence of hyphae in cortical cells (cc) (is, intercellular spaces; n, orchid cell nucleus); (D) magnification detail of infected cells with view of elongated fungal pelotons, cut transversally in P1 and longitudinally in P2 (C and D optical microscopy stained with Trypan Blue); (E) TEM of hyphae (h) in a peloton cut transversally (as in P1), with a dolipore (d) in one hypha; (F) TEM view of a dolipore with parenthesome (p) in a fungal hypha; (G) TEM detail of a cell with longitudinal cut of the peloton (as in P2), showing a clamp connection (c-c). Note the host membrane surrounding the hyphae (arrows in E and G). Scale bars: (A, B) ¼ 0.5 cm; (C, D) ¼ 50 mm; (E) ¼ 2 mm; (F) ¼ 1 mm; (G) ¼ 2 mm. sequences of ascomycetes, more rarely zygomycetes or even a Inocybe occurred exclusively in 75 % of plants; it was thus basidiomycete ( Thelephora , Table 1). With the exception of the most abundant fungus across the range and at the plant Thelephora , an ECM genus, these additional fungi were scale (detected in 78 % of root fragments per plant on either soil saprobes ( Paecilomyces and Metarhizium species) average). Identical Inocybe -related ITS sequences (or .97 % or parasites and possible endophytes ( Neonectria , Didymella , similar, when considering cloned sequences) were most Olpidium and Protoventuria species; see Table 1). Two pro- often retrieved from the same plant (e.g. two fragments produced only ascomycetes (EM10 and EM11 from Saint duced the same ITS in NA1; Table 1). At the population Cle´ment; Table 1). level, only one sequence was shared between two plants Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species Page 9 of 16

A B

G F

F I G . 4. Morphology and absence of fungal colonization of stolons and lateral bulbils (b): (A) apex of a stolon with bulbils and scaly sheathing leaf (l) and vascular bundle (vb) in transparent stolon tissues; (B) detail of the sheathing leaf protecting a bulbil; (C) sheathing leaf pushed aside to show rhizoids (r) covering a bulbil; (D) transverse section of a stolon with large, empty cells and its lateral bulbil with starch-ﬁlled cells (bc); (E) contact between stolon and bulbil (n, cell nucleus); (F) longitudinal section showing the smaller cells in bulbils; (G) detailed view of the meristematic zone (m) at the bulbil apex. (D–G) Optical microscopy stained with safranin O and fast green FCF. Scale bars: (A–C) ¼ 1 mm; (D–G) ¼ 100 mm.

(NB1 and NC1). Conversely, divergent sequences were often I. subnudipes , I. glabripes , I. dulcamara and I. terrigena ), found in single plants (up to 11 sequences in EM20; Table 1). whereas clades II and V remained unidentified. Sequences In an analysis of Inocybe , the various sequences retrieved obtained from the same plant often clustered together; in clustered into seven well-supported clades (Fig. 5), suggesting three plants (EM20, EM26 and EM27), sequences clustered that E. aphyllum is not highly specific at the intrageneric level. in the closely related clades I and II (Fig. 5). In well-sampled Five clades clustered with identified species (two-thirds of populations, such as at Saint Cle´ment or Thue`s-entre-Valls, the sequences clustered with I. fuscidula , others with sequences from different clades were retrieved (Fig. 5), Page 10 of 16 Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species

EU711163 Luz Saint Sauveur EU711225 Picherande France EU711224 Picherande France EU711226 Picherande France EU711223 Picherande France EU711219 Picherande France EU711205 Saint Paul de Salers France EU711195 Oo France EU711198 Oo France EU711196 Oo France EU711203 Saint Paul de Salers France EU711215 Picherande France EU711213 Picherande France 80 EU711208 Saint Paul de Salers France EU711207 Saint Paul de Salers France EU711221 Picherande France 89 EU711214 Picherande France EU711209 Thues entre valls France 88 EU711199 Oo France EU711197 Oo France Clade I EU711222 Picherande France EU711202 Saint Paul de Salers France EU711206 Saint Paul de Salers France EU711190 Cazaux France EU711189 Cazaux France EU711187 Cazaux France EU711210 Thues entre valls France EU711204 Saint Paul de Salers France EU711234 Nagano Japan 100 EU711235 Nagano Japan EU711233 Nagano Japan EU711236 Nagano Japan EU711237 Nagano Japan EU711238 Nagano Japan EU711217 Picherande France 100 EU711188 Cazaux France Inocybe fuscidula 1 Inocybe fuscidula 5 Inocybe fuscidula 6 Inocybe fuscidula 4 Inocybe fuscidula 3 EU722336 Cazaux France EU711211 Thues entre valls France EU711185 Saint Clement France EU722337 Cazaux France Inocybe flocculosa 3 100 Inocybe flocculosa 2 100 Inocybe flocculosa 4 99 Inocybe flocculosa 1 Inocybe splendens Inocybe fuscidula var bisporigera

F I G . 5. Phylogenetic position of the E. aphyllum mycorrhizal fungi related to other Inocybe species (including the subgenus Mallocybe ) and Hebeloma (ML on an alignment of the ITS þ 28S rDNA; GTR model, 1000 bootstrap; only bootstraps .80 % are shown). Russian samples for which fungal ITSs were not sequenced are excluded. further supporting a low speciﬁcity within Inocybe . With the geographic areas (Fig. 5). A Mantel test showed that similarity exception of clades III (from Japan only) and VII (a single between Inocybe -related sequences and distance was positively sequence from France), each clade was found over large correlated ( R ¼ 0.174; P ¼ 0.007); more distant plants had Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species Page 11 of 16

Clade I EU711194 Oo France EU711193 Oo France EU711192 Oo France EU711201 Oo France 100 EU711200 Oo France Clade II EU711216 Picherande France EU711191 Oo France EU711220 Picherande France 100 Inocybe nitidiuscula Inocybe rufuloides EU711176 Thues entre valls France 100 Inocybe geophylla 2 95 Inocybe geophylla 1 Inocybe armeniaca EU711242 Yamanashi Japan 91 EU711241 Yamanashi Japan Clade III 100 EU711243 Yamanashi Japan Inocybe subnudipes Inocybe glabripes 2 98 Inocybe glabripes 1 Clade IV EU711175 Thues entre valls France 100 Inocybe griseolilacina 1 100 Inocybe griseolilacina 2 EU711172 Lans en Vercors France 100 EU711171 Lans en Vercors France EU711170 Lans en Vercors France 100 Clade V 96 EU711184 Saint Clement France 100 EU711183 Ural Russia EU711166 Ponteils France 94 EU711164 Ponteils France 94 EU711182 Saint Clement France Inocybe dulcamara 5 83 Inocybe dulcamara 2 Clade VI 100 Inocybe dulcamara 4 95 Inocybe substraminipes Inocybe dulcamara 3 100 Inocybe terrigena EU711173 Villard de Lans France Clade VII 97 Inocybe dulcamara 1 Mallocybe unicolor 85 Inocybe mixtilis Inocybe calospora 96 Inocybe curvipes Inocybe stellatospora 100 Inocybe geophylla 5 97 Inocybe geophylla 4 Inocybe geophylla 3 Inocybe relicina 100 Inocybe lanuginosa 1 97 Inocybe calamistrata 90 Inocybe whitei Inocybe rufofusca Inocybe lanuginosa 2 Inocybe lacera Inocybe tubarioides EU711240 Yamanashi Japan EU711239 Yamanashi Japan 91 Hebeloma leucosarx Hebeloma velutipes 97 Hebeloma pusillum 99 Hebeloma mesophaeum Galerina autumnalis Cortinarius odoratus

200·0

F I G . 5. Continued. Page 12 of 16 Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species more similar fungal ITS. However, considering data from N. nidus-avis is the scarcity of hyphae directly linking root France only, the correlation was significantly negative ( R ¼ tissues and soil (Selosse et al. , 2002 b; Selosse, 2003). –0 .226; P ¼ 0.0001). On average, similarity between diver- There is a sharp separation between infection and storage gent Inocybe ITS sequences from the same plant (61 + 21 zone, i.e. inflorescence buds, which may act as reserves for %, mean + standard deviation) or the same population (66 flowering (Rasmussen, 1995), and bulbils. MH orchids show % + 24 %) was significantly higher than between ITS from variable modalities of separation between storage cells and different populations (53 + 19 %, P , 0.0001 in both cases), colonized cells, e.g. between cells from the same tissue (e.g. suggesting a geographical structure at smaller scale. No sig- in Corallorhiza trifida ; Scrugli et al. , 1995), for which the nificant correlation with elevation was found (not shown). adaptative value is unclear. Xylem was poorly developed (and only limited lignification was observed; Fig. 4), a feature common to many MH plants (Leake, 1994) for Investigation of surrounding ECM tips which phloem distributes nutrients. Correlatively, inflores- Since Inocybe form ECMs, to find them a search was carried cences need wet springs (see Rasmussen, 1995) and appear out on tree roots around two plants of the Saint Cle´ment popu- water-filled, a feature suggesting that turgidity contributes to lation (Table 1). The 71 ECM tips collected around orchid the erect habit. Another fluid transport system, the aerench- EM15 exhibited 25 ITS RFLP patterns, one of which was iden- yma, was present in the form of large intercellular spaces tical to that obtained from EM15 (not shown). Its sequence was (Fig. 3C). It explains the presence of stomata on the under- identical to that of the orchid (an Inocybe ITS, EU711184). It ground rhizome shoots (Leake, 1994; our personal obser- colonized eight tips (11 .2 %), and was the second by order of vations), a somewhat paradoxical feature since they are abundance in this ECM sampling. The 46 ECM tips collected usually involved in CO 2 uptake and are therefore absent around EM12 exhibited ten ITS RFLP patterns, but none of from most MH plants (Leake, 1994). Intercellular spaces are these matched that from the neighbouring orchid. not devoted to hyphal spread in rhizomes (Fig. 3C, D) but rather may be linked to survival of E. aphyllum in wet soils (e.g. the observed Saint Cle´ment population) and valleys DISCUSSION near rivers (e.g. populations Ooˆ, Cazaux, Picherande, Ponteils, Thue`s-entre-Valls and Nagano Prefecture) by allow- Underground morphology of E. aphyllum ing circulation of oxygen in submerged conditions. Compared with other orchids, E. aphyllum has an unusual, complex subterranean structure (Figs 1, 3 and 4); such a Asexual reproduction in E. aphyllum highly derived morphology is common for MH orchids (Ramussen, 1995) and other MH plants (Leake, 1994; In contrast to the coralloid rhizomes, the thin stolons Imhoff, 2003; Klimesˇova´, 2007). Absence of roots is a (¼ runners) radiate away from the main rhizome and bear shared feature with MH Corallorhiza species (Fu¨ller, 1977; starch-filled bulbils, probably involved in asexual reproduction Ramussen, 1995). The term ‘mycorrhizal’ fungus is used (Fig. 4). Bulbils are surrounded by rhizoids and thus reminis- here in the enlarged meaning of ‘an underground fungal cent of protocorms (Fig. 4), a heterotrophic stage in orchid associate, having a nutritional role’. As expected from the lit- seed germination. They may reiterate the same developmental erature (Irmisch, 1853; Ziegenspeck, 1936), the plant encom- process (Fig. 1), except for the fact that they are initially apos- passes two kinds of specialized underground shoots: thin ymbiotic (see below). MH plants have often evolved under- stolons and plagiotropic coralloid rhizomes. Since numerous ground asexual multiplication (Leake, 1994; Selosse, 2003), inflorescence buds were found on some flowering plants, the although the adaptative basis for this is unclear (Klimesˇova´, common idea that ramets die after fruiting (monocarpic devel- 2007). Their development into adult plants was did not opment, e.g. Ziegenspeck, 1936; Rasmussen, 1995) is ques- observed, but the existence of similar bulbils in E. roseum tionable. Instead, periods of underground growth could that develop into new rhizomes (Yagame et al. , 2007), together explain the irregular appearance of inflorescences for each with reserves and a meristem, makes them candidates for plant (Summerhayes, 1951; Soyrinki, 1987; Robin, 1999). asexual reproduction; accordingly, Rasmussen (1995) indi- Ramet survival thus deserves further studies. cated that they develop into ‘small plants’ during the Coralloid rhizomes are densely branched (Figs 1 and 3) autumn. Epipogium aphyllum forms few fruits (Fu¨ller, 1977; since most axillary buds produce ramifications (Ziegenspeck, Van der Cingel, 1995; Harrap and Harrap, 2005; our personal 1936), and drastically differ from the simple tuberous observations); despite their strong vanilla smell, flowers are rhizome found in E. roseum and some other Epipogium poorly visited, possibly due to a combination of low nectar species (Pridgeon et al. , 2005; Yamato et al. , 2005). production and rarity of insects in the dark habitats of Coralloid rhizomes are densely colonized by fungi (Fig. 3) E. aphyllum (Vo¨th, 1994). Since animals often eat inflores- and therefore constitute a nutritional organ. Fungal coloniza- cences (slugs and deer: Harrap and Harrap, 2005; tion closely resembles that in other orchids (Scannerini and Kopylov-Gus’kov et al. , 2007; our personal observations), Bonfante, 1983; Smith and Read, 1997), with a final lysis of sexual reproduction seems of low efficiency. It is unclear pelotons (not shown) and no colonization of meristems, inflor- whether this facilitated an increase in asexual reproduction escence buds and central cylinder. Lack of colonization of or, at the other extreme, low fruit set could be sustained outer rhizome cells (Fig. 3C) is convergently reported from after establishment of an efficient asexual reproduction. the MH orchid Neottia nidus-avis (Selosse, 2003) and some Nevertheless, asexual reproduction accounts for the existence other MH plants (Leake, 1994). Another similarity with of groups of ramets (Kopylov-Gus’kov et al. , 2007) and for Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species Page 13 of 16 the observation that populations often extend downward along 2005). Direct amplifications of Xerocomus and Lactarius valleys (Robin, 1999), as expected if gravity or water disperse species suggest that they are common in investigated rhizomes, bulbils after disturbance. but their exact status remains unknown. Similarly, an ECM It is noteworthy that both molecular and microscopic inves- Thelephora species was found when cloning EM5. Such tigations failed to detect fungal colonization in stolons and ECM genera are not usual contaminants. Thus, although bulbils, so the fungus is probably independently transmitted. Inocybe -related sequences were recovered from the same Lack of fungal colonization was also reported in E. roseum populations, it is not possible to rule out that these ECM stolons (Yamato et al. , 2005). Epipogium aphyllum bulbils species are truly mycorrhizal. have rhizoids that were not described for E. roseum and may The identification of Inocybe symbionts is congruent with represent entry points for fungi, as described in protocorms the presence of dolipores with perforate parenthesomes (Rasmussen, 1995) and adults of E. aphyllum (Scrugli et al. , (Fig. 3F) and clamp connections (Fig. 3G) on intracellular 1995). In some MH plants, symbionts are directly transmitted hyphae; clamp connections were also reported by Scrugli during underground asexual multiplication, such as in root pro- et al. (1995) and Rasmussen (1995). Based on peloton mor- pagules of Arachnitis uniflora (Domıńguez et al. , 2006) or in phology, Scrugli et al. (1995) described two kinds of fungi, roots separated from rhizomes that develop into new shoots each in different cell layers. It is unclear whether they rep- in Neottia nidus-avis (Selosse, 2003). resent different species (perhaps explaining the diverse A main difference between asexual reproduction in sequences recovered from some individuals) or different devel- Epipogium species and other MH plants is the distance from opmental stages of a single Inocybe symbiont. Given the the mother plant, which is due to stolons. This could mean results of direct peloton analysis, we favour the later alterna- that ( a) the fungus would be difficult and perhaps costly to tive. To our knowledge, no MH orchids were hitherto reported maintain in long, rapidly growing stolons and ( b) fungal pre- to associate with Inocybe. Inocybe species have been reported sence in bulbils would not ensure that the fungus will form from some partly heterotrophic orchids ( Epipactis and ECM around bulbils (the required ultimate carbon source, Cephalanthera ; Bidartondo et al. , 2004), but no evidence see below). Inocybe genets can be ,2.5 m in diameter was obtained that they actually formed pelotons. Inocybe (Lilleskov et al. , 2004), and ramets are likely to be smaller. was one of the rare large groups of ECM fungi not shown to It is thus proposed that the ability to produce bulbils meant have been recruited by MH orchids. that no direct transmission was selected in E. aphyllum . Epipogium aphyllum associates with a great range of Accordingly, the aposymbiotic state of E. aphyllum propagules Inocybe species. Interspecific ITS divergence is considered correlates to the diversity of fungi retrieved from different to be at least 3 % within Inocybe species (Matheny et al. , plants in a population (Table 1; see below) because indepen- 2005); based on this threshold, then at least 22 species were dent transmission potentially allows each new ramet to associ- probably encountered here. The exact range of associated ate with a different fungus. The diverse structures involved in fungi in E. roseum is unknown because only three Japanese underground sprouting of MH plants, showing variable fungal populations of this species were investigated (Yamato et al. , transmission, probably result from independent evolutionary 2005). There is no clear geographic pattern of association adaptations of existing structures, with variable trade-offs with Inocybe or support for geographical E. aphyllum races between exploitation of the fungi and dissemination. differing in their fungal partners (Figs 2 and 5). However, the sampling in the present study poorly covers the Eurasian range, because it was not possible to obtain permits to Fungal associates of E. aphyllum across Eurasia collect in some countries. Although conceived for plant pro- The most commonly identified symbionts belonged to tection, such limitations hinder cross-border movement of Inocybe subgenera Mallocybe and Inocybe sensu stricto ; this scientific samples and thus biological knowledge for protected is a common and worldwide genus of ECM fungi, present species such as orchids (Roberts and Solow, 2008). The diver- throughout the Eurasian range of E. aphyllum (Matheny sity of associated Inocybe species could be explained by the et al. , 2005; Ryberg et al. , 2008). In two cases, only ascomy- existence, even in sympatry, of several cryptic species differ- cetes were recovered in cloning procedures that provided two ing in Inocybe preference. Given the importance of asexual clones each (EM10 and EM11, Table 1). Among all clones, reproduction of E. aphyllum , as previously discussed, emer- the average probability of encountering ascomycetes among gence of local races with diverging specificities would be poss- clones was 0 .33 (thus, P ¼ 0.33 2 ¼ 0.11 for two clones, as ible. Subspecies with different fungal preferences among in EM10 and EM11, assuming a constant probability). This Sebacinales were reported in the MH Hexalectris spicata high probability, together with the possible endophytic or (Taylor et al. , 2003), and cryptic species differing in associated saprophytic ecology of these fungi, does not support the Russulaceae exist in the MH Corallorhiza maculata (Taylor hypothesis that ascomycetes were the sole mycorrhizal fungi. et al. , 2004). Although the great diversity of E. aphyllum Additionally, they were not seen in TEM investigations. ITS sequences (Fig. 1) may reflect speciation, the present Similar ascomycete taxa had already been recovered from data rather suggest a single species with a low specificity: other orchids when cloning fungal ITS (e.g. Julou et al. , first, some individuals harbour different partners (e.g. EM20, 2005; Abadie et al. , 2006). Conversely, in four occurrences, 26 or 27; Table 1), supporting the low specificity within ITS sequences of Hebeloma , Xerocomus and Lactarius Inocybe ; and secondly, individuals from the same population species were directly amplified (Table 1). Hebeloma sequences differ in fungal associates (e.g. Saint Cle´ment or were amplified from pelotons, making them likely symbionts Thue`s-entre-Valls; Table 1), although they are likely to result (Hebeloma and Inocybe are closely related; Matheny et al. , from asexual reproduction (see above). Page 14 of 16 Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species

This work has consequences for conservation of ACKNOWLEDGEMENTS E. aphyllum . First, since Inocybe species are not now cultur- We thank Jean-François Bergouignan, Christine Casiez, Laure able in vitro (Matheny et al. , 2005), ex situ conservation and Civeyrel, Edouard Chas, Herve´ Christophe, Jean Dauge, germination are impossible. However, since some Hebeloma Ge´rard Joseph, Jean Marc Lewin, Tastuya Nagai, Yuzan species are culturable, their ability to germinate E. aphyllum Okuyama, Re´my Souche and Irina Tatarenko for help in requires further studies. Secondly, the present data show sampling as well as Marie-Pierre Dubois, Pierre-Henri Fabre, that Inocybe symbionts form ECM with nearby trees, Prune Pellet and Annie Tillier for help in molecular and phy- making them the most probable carbon source. Thus, trees logenetic studies. Thierry Vindolet provided crucial help with should be protected around E. aphyllum populations. Since light microscopy investigations. We thank Mark Chase and Inocybe species are generally non-specific ECM associates two anonymous referees for excellent suggestions on an of trees (Ryberg et al. , 2008), there are no obvious tree earlier version of this article. M.-A. Selosse is funded by the species to be favoured. Asexual multiplication suggests Centre National de la Recherche Scientifique and the Socie´te´ that soil disturbance may contribute to dispersal at a local Française d’Orchidophilie. P. Bonfante is funded by the scale. Finally, the present observations confirm the Biodiversity Project, Commessa ‘Istituto per la Protezione existence of large underground rhizomes described by other delle Piante’ (CNR, Italy). authors (e.g. Rasmussen, 1995; Kopylov-Gus’kov et al. , 2007), so that extinction of populations cannot be assessed by observations of inflorescences only (Harrap and Harrap, 2005). LITERATURE CITED Abadie JC, Pu¨ttsepp U ¨ , Gebauer G, Faccio A, Bonfante P, Selosse MA. Evolution of fungal associations in MH orchids 2006. Cephalanthera longifolia (Neottieae, Orchidaceae) is mixotrophic: a comparative study between green and non-photosynthetic individuals. Although a few additional and perhaps questionable sym- Canadian Journal of Botany 84 : 1462–1477. bionts may occur, E. aphyllum appears to associate specifically Bidartondo MI. 2005. The evolutionary ecology of mycoheterotrophy. New with a single ECM clade. This fits the usual paradigm for MH Phytologist 167 : 335–352. plants from temperate regions (Taylor et al. , 2002; references Bidartondo MI, Burghardt B, Gebauer G, Bruns TD, Read DJ. 2004. Changing partners in the dark: isotopic and molecular evidence of ecto- in the Introduction). It contrasts with the MH sister species mycorrhizal liaisons between forest orchids and trees. Proceedings of E. roseum (Fig. 2) convincingly reported to associate with the Royal Society of London, Series B, Biological Sciences 271 : saprobic Coprinaceae (Yamato et al. , 2005; Yagame et al. , 1799–1806. 2007). The host jump is not unexpected in itself, as it has Bougoure JJ, Dearnaley JDW. 2005. The fungal endophytes of Dipodium variegatum (Orchidaceae). Australasian Mycologist 24 :15–19. been demonstrated among MH orchids such as Corallorhiza Bryan JH. 1955. Differential staining with a mixture of safranin and fast green (Taylor and Bruns, 1997, 1999; Taylor et al. , 2004) and MH FCF. Staining Technology 30 : 153–157. Ericaceae (Bidartondo, 2005). Host jumps are often associated Cameron DD, Leake JR, Read DJ. 2006. Mutualistic mycorrhiza in orchids: with speciation in MH plants (see also cryptic species evidence from plant–fungus carbon and nitrogen transfers in the green- described above), although currently it is unknown whether leaved terrestrial orchid Goodyera repens . New Phytologist 171 : 405–416. this is a cause or a consequence of speciation. Campbell EO. 1970. Morphology of the fungal association in three species of The change in ecology of associated fungi is, however, an Corallorhiza in Michigan. Michigan Botanist 9: 108–113. unexpected feature, and deserves further study in other Cha JY, Igarashi T. 1996. Armillaria jezoensis , a new symbiont of Galeola Epipogium species because it raises many important questions. septentrionalis (Orchidaceae) in Hokkaido. Mycoscience 37 : 21–24. Danton P, Baffray M. 2005. Inventaire des plantes prote´geés en France. Paris: For example, E. roseum has a simpler rhizome morphology Nathan Ed. and a faster life cycle [ ,1 year (Yagame et al. , 2007) Dearnaley JDW. 2006. The fungal endophytes of Erythrorchis cassythoides — versus 10 years for E. aphyllum (Irmisch, 1853; Rasmussen, is this orchid saprophytic or parasitic? Australasian Mycologist 25 : 1995)]. Is this an adaptation to saprobic fungi, perhaps 51–57. because they persist less well or are active over shorter Dearnaley JWD. 2007. Further advances in orchid mycorrhizal research. Mycorrhiza 17 : 475–486. periods than ECM fungi? Moreover, E. roseum is tropical, as Dearnaley JDW, Le Brocque AF. 2006. Molecular identification of the is Eulophia zollingeri which associates with the same saprobic primary root fungal endophytes of Dipodium hamiltonianum (yellow hya- fungal clade (Ogura-Tsujita and Yukawa, 2008). Other orchids cinth orchid). Australian Journal of Botany 54 : 487–491. putatively associated with saprobic fungi also tend to be Domıńguez L, Se´rsic A, Melville L, Peterson RL. 2006. ‘Prepackaged symbioses’: propagules on roots of the myco-heterotrophic plant Arachnitis tropical (Terashita, 1985; Lan et al. , 1994; Umata et al. , uniflora . New Phytologist 169 : 191–198. 1995; Cha and Igarashi, 1996; Dearnaley, 2006). It is probable Environment Agency of Japan . 2000. Threatened wildlife of Japan, red data that higher decay rates due to hot and wet tropical climates book , 2nd edn. Vol. 8. Vascular plants . Tokyo: Japan Wildlife Research make saprobic fungi better able to obtain carbon and support Center. growth of MH plants. Epipogium aphyllum also occurs in tro- Fu¨ller F. 1977. Limodorum, Epipogium, Neottia, Corallorhiza. Wittenberg Lytherstadt: A. Ziemsen Verlag. pical regions (Pridgeon et al. , 2005); although some ECM Girlanda M, Selosse MA, Cafasso D, et al. 2006. Inefficient photosynthesis fungi also occur in these regions, the fungal associates of in the Mediterranean orchid Limodorum abortivum is mirrored by specific such plants, if conspecific, deserve further study. To address association to ectomycorrhizal Russulaceae. Molecular Ecology 15 : these issues and better understand MH biology on a global 491–504. Hall TA. 1999. BioEdit: a user-friendly biological sequence alignment editor scale, investigations on other tropical MH orchids and com- and analysis program for Windows 95/98/NT. Nucleic Acids Symposium parison with related MH species from temperate regions are Series 41 : 95–98. now required. Harrap A, Harrap S. 2005. Orchids of Britain and Ireland . London: C. Helm. Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species Page 15 of 16

Hoch HC. 1986. Freeze-substitution of fungi. In: Aldrich HC, Todd WJ, eds. Robin M. 1999. Evolution d’une population d’ Epipogium aphyllum : une Ultrastructure techniques of microorganisms , New York, NT: Plenum eńigme de´voileé ? Cahiers de la Socie´te´ Française d’Orchidophilie 6: Press, 183–211. 254–260. Imhof S. 2003. A dorsiventral mycorrhizal root in the achlorophyllous Rodriguez R, Oliver JL, Marin A, Medina JR. 1990. The general stochastic Sciaphila polygyna (Triuridaceae). Mycorrhiza 13 : 327–332. model of nucleotide substitution. Journal of Theoretical Biology 142 : Irmisch T. 1853. Beitra¨ge zur Morphologie und Biologie der Orchideen . 485–501. Leipzig. Ryberg M, Nilsson RH, Kristiansson E, To¨pel M, Jacobsson S, Larsson E. Julou T, Burhardt B, Gebauer G, Berviller D, Damesin C, Selosse MA. 2008. Mining metadata from unidentified ITS sequences in GenBank: a 2005. Mixotrophy in orchids: insights from a comparative study of case study in Inocybe (Basidiomycota). BMC Evolutionary Biology 8: 50. green individuals and non-photosynthetic mutants of Cephalanthera Saitou N, Nei M. 1987. The neighbor-joining method: a new method for damasonium . New Phytologist 166 : 639–653. reconstructing phylogenetic trees. Molecular Biology and Evolution 4: Klimesˇova´ J. 2007. Root-sprouting in myco-heterotrophic plants: prepackaged 406–425. symbioses or overcoming meristem limitation? New Phytologist 173 : Scannerini S, Bonfante P. 1983. Comparative ultrastructural analysis of 8–10. mycorrhizal associations. Canadian Journal of Botany 61 : 917–943. Kopylov-Gus’kov YO, Vinogradova TN, Lyskov DF, Volkova PA. 2007. Scrugli A, Cogoni A, Riess S. 1995. Mycorrhizal endophytes of the Population of Epipogium aphyllum (F.W. Schmidt) Sw., Orchidaceae chlorophyll-free orchids, Corallorhiza trifida Chatelain and Epipogium Juss.) on the Kem’-Ludskij archipelago (White Sea). In: Tverskoj gosu- aphyllum Swartz, under light microscopy and true confocal scanner darstvennyj universitet , Tver. microscopy. Caesiana 5: 29–38. Koske RE, Gemma JN. 1989. A modified procedure for staining roots to Selosse MA. 2003. La neóttie, une “mangeuse” d’arbres. L’Orchidophile 155 : detect VA mycorrhizas. Mycological Research 92 : 486–505. 21–31. Kull T, Hutchings MJ. 2006. A comparative analysis of decline in the distri- Selosse MA, Bauer R, Moyersoen B. 2002 a. Basal hymenomycetes belong- bution ranges of orchid species in Estonia and the United Kingdom. ing to the Sebacinaceae are ectomycorrhizal on temperate deciduous trees Biological Conservation 129 : 31–39. in silva: microscopic and molecular evidence. New Phytologist 155 : Lan J, Xu JT, Li JS. 1994. Study on symbiotic relation between Gastrodia 183–195. elata and Armillariella mellea by autoradiography. Acta Mycologica Selosse MA, Weiß M, Jany JL, Tillier A. 2002 b. Communities and popu- Sinica 13 : 219–222. lations of sebacinoid basidiomycetes associated with the achlorophyllous Lanave C, Preparata G, Saccone C, Serio G. 1984. A new method for cal- orchid Neottia nidus-avis (L.) L.C.M. Rich. and neighbouring tree ecto- culating evolutionary substitution rates. Journal of Molecular Evolution mycorrhizae. Molecular Ecology 11 : 1831–1844. 20 : 86–93. Selosse MA, Faccio A, Scappaticci G, Bonfante P. 2004. Chlorophyllous and Leake JR. 1994. The biology of myco-heterotrophic (‘saprophytic’) plants. achlorophyllous specimens of Epipactis microphylla (Neottieae, Orchidaceae) are associated with ectomycorrhizal septomycetes, includ- New Phytologist 127 : 171–216. ing truffles. Microbial Ecology 47 : 416–426. Leake JR. 2004. Myco-heterotroph/epiparasitic plant interactions with ecto- Smith SE, Read DJ. 1997. Mycorrhizal symbiosis , 2nd edn. London: mycorrhizal and arbuscular mycorrhizal fungi. Current Opinion in Academic Press. Plant Biology 7: 422–428. Soyrinki N. 1987. On the periodicity in the flowering of Epipogium aphyllum Lilleskov EA, Bruns TD, Horton TR, Taylor DL, Grogan P. 2004. (Orchidaceae). Memoranda Societas Pro Fauna et Flora Fennica 63 : Detection of forest stand-level spatial structure in ectomycorrhizal 63–72. fungal communities. FEMS Microbiology Ecology 49 : 319–332. Summerhayes VS. 1951. Wild orchids of Britain with a key to the species . McKendrick SL, Leake DJ, Taylor DL, Read DJ. 2000. Symbiotic germina- London: Collins. tion and development of myco-heterotrophic plants in nature: ontogeny of Swofford DL. 2004. PAUP*: Phylogenetic analysis using parsimony (*and Corallorhiza trifida and characterization of its mycorrhizal fungi. New other methods), version 4.0 . Sunderland, MA: Sinauer. Phytologist 145 : 523–537. Taylor DL, Bruns TD. 1997. Independent, specialized invasions of ectomy- McKendrick SL, Leake JR, Taylor DL, Read DJ. 2002. Symbiotic germina- corrhizal mutualism by two non photosynthetic orchids. Proceedings of tion and development of the myco-heterotrophic orchid Neottia nidus-avis the National Academy of Sciences of the USA 94 : 4510–4515. in nature and its requirement for locally distributed Sebacina spp. New Taylor DL, Bruns TD. 1999. Population, habitat and genetic correlates of Phytologist 154 : 233–247. mycorrhizal specialization in the cheating orchids Corallorhiza maculata Maekawa F. 1971. The wild orchids of Japan in color . Tokyo: and C. mertensiana . Molecular Ecology 8: 1719–1732. Seibundo-shinkosha [in Japanese with English summary]. Taylor DL, Bruns TD, Leake JR, Read D. 2002. Mycorrhizal specificity and Matheny PB. 2005. Improving phylogenetic inference of mushrooms function in myco-heterotrophic plants. In: Van der Heijden MGA, with RPB1 and RPB2 nucleotide sequences (Inocybe, Agaricales). Sanders I, eds. Mycorrhizal ecology . Berlin: Springer Verlag, 375–413. Molecular Phylogenetics and Evolution 35 : 1–20. Taylor DL, Bruns TD, Szaro TM, Hodges SA. 2003. Divergence in mycor- Molvray M, Kores PJ, Chase MW. 2000. Polyphyly of mycoheterotrophic rhizal specialization within Hexalectris spicata (Orchidaceae), a nonpho- orchids and functional influences on floral and molecular characters. In: tosynthetic desert orchid. American Journal of Botany 90 : 1168–1179. Wilson KL, Morrison DA, eds. Monocots: systematics and evolution . Taylor DL, Bruns TD, Hodges SA. 2004. Evidence for mycorrhizal races in a Melbourne: CSIRO, 441–447. cheating orchid. Proceedings of the Royal Society of London, Series B, Ogura-Tsujita Y, Yukawa T. 2008. High mycorrhizal specificity in a wide- Biological Sciences 271 : 35–43. spread mycoheterotrophic plant, Eulophia zollingeri (Orhcidaceae). Terashita T. 1985. Fungi inhabiting orchids in Japan (III). A symbiotic exper- American Journal of Botany 95 : 93–97. iment with Armillaria mellea and Galeola septentrionalis . Transactions Posada D, Crandall KA. 1998. Modeltest: testing the model of DNA substi- of the Mycological Society of Japan 26 : 47–53. tution. Bioinformatics 14 : 817–818. Thompson JD, Higgins DG, Gibson TG. 1994. CLUSTAL W: improving the Pridgeon AM, Cribb PJ, Chase MW, Rasmussen FN. 2005. Genera orchi- sensitivity of progressive multiple sequence alignment through sequence dacearum . Vol. 4. Epidendroidea (part one). Oxford: Oxford University weighting, position-specific gap penalties and weight matrix choice. Press. Nucleic Acids Research 22 : 4673–4680. Rasmussen HN. 1995. Terrestrial orchids – from seed to mycotrophic plant . Umata H. 1995. Seed germination of Galeola altissima , an achlorophyllous Cambridge: Cambridge University Press. orchid, with aphyllophorales fungi. Mycoscience 36 : 369–372. Rasmussen HN. 2002. Recent developments in the study of orchid mycor- Van der Cingel NA. 1995. An atlas of orchid pollination, European orchids . rhiza. Plant and Soil 244 : 149–163. Rotterdam: A. A. Balkema. Reynolds EW. 1963. The use of lead citrate at high pH as an electron opaque Vilgalys R, Hester M. 1990. Rapid genetic identification and mapping of stain in electron microscopy. Journal of Cellular Biology 17 : 208–212. enzymatically amplified ribosomal DNA from several Cryptococcus Roberts DL, Solow AR. 2008. The effect of the convention on international species. Journal of Bacteriology 172 : 4238–4246. trade in endangered species on scientific collections. Proceedings of the Vo¨th W. 1994. Sind Blu¨ten von Epipogium aphyllum Sw. entomogam oder Royal Society 275 : 987–989. autogam? Die Orchideen 45 : 248–251. Page 16 of 16 Roy et al. — Epipogium aphyllum associates with ectomycorrhizal Inocybe species

Warcup JH. 1991. The Rhizoctonia endophytes of Rhizanthella Yagame T, Yamato M, Suzuki A, Iwase K. 2008. Ceratobasidiaceae mycor- (Orchidaceae). Mycological Research 95 : 656–664. rhizal fungi isolated from nonphotosynthetic orchid Chamaegastrodia Warcup JH, Talbot PHB. 1967. Perfect states of rhizoctonias associated with sikokiana . Mycorrhiza 18 : 97–101. orchids. New Phytologist 66 : 631–641. Yamato M, Iwase K. 2008. Introduction of asymbiotically propagated seed- Weiss M, Selosse MA, Rexer KH, Urban A, Oberwinkler F. 2004. lings of Cephalanthera falcata (Orchidaceae) into natural habitat and Sebacinales: a hitherto overlooked cosm of heterobasidiomycetes investigation of colonized mycorrhizal fungi. Ecological Research 23 : with a broad mycorrhizal potential. Mycological Research 108 : 329–337. 1003–1010. Yamato M, Yagame T, Suzuki A, Iwase K. 2005. Isolation and identification Yagame T, Yamato M, Mii M, Suzuki A, Iwase K. 2007. Developmental of mycorrhizal fungi associating with an achlorophyllous plant, process of an achlorophyllous orchid, Epipogium roseum Epipogium roseum (Orchidaceae). Mycoscience 46 : 73–77. (D.Don) Lindl. from seed germination to flowering under symbiotic Ziegenspeck H. 1936. Orchidaceae. In: Wangerin W, Schro¨ter C, eds. cultivation with a mycorrhizal fungus. Journal of Plant Research 120 : Lebensgeschichte der Blu¨tenpflanzen Mitteleuropas , Vol. 4. Stuttgart: 229–236. Eugen Ulmer, 1–840. 79.(1*= 79.(1*= = = _= `= _= `= _= `= _= `= _= `= _= :_= _= >(4-*9*749745-.(=47(-.)8=+742=-&.1&3)=9745.(&1=+47*898=&884(.&9*=<.9-=&='74&)=).;*78.9>= 4+=*(942>(477-.?&1=+:3,. _=:'2.99*)=94== = &=51:5&79=)*8=2>(4-S9S749745-*8=8439=57S8*39*8=*3=8.*=):=:)8 89=2&.8=97T8=5*:=>= 439=S9S=S9:).S*8_= 1=7*88479=)*8=S9:)*8=57S(S)*39*8=6:*=1*8=2>(4-S9S749745-*8=9*25S7S*8=8439= &884(.S*8= 85S(.+.6:*2*39= A= )*8= (-&25.,3438= *(942>(47-.?.*38= 9&3).8= 6:*= 1*8= 2>(4-S9S749745-*8= 9745.(&1*8= 8439= &884(.S*8= 85S(.+.6:*2*39= A= )*8= (-&25.,3438= 8&5745->9*8_= S9:)*= )*= _= &5->11:2 = (43+.72*= (*99*= 9*3)&3(*= *9= 7S;T1*= 6:*= )*8= *85T(*8= 574(-*8= 5*:;*39= U97*= &884(.S*8= A= )*8= (-&25.,3438= 97T8= ).++S7*398`= 9&39= ):= 54.39= )*= ;:*= S(414,.6:*= 6:*= 5->14,S3S9.6:*_= .= )*= 9*11*8= ).++S7*3(*8= 5*:;*39= U97*= &997.':S*8= &:== (-&3,*2*398= )*= (422:3&:9S8= )*= (&-25.,3438`= 43= 5*:9= 8*= )*2&3)*7= (*= 6:.1= *3= *89= )*8= 2>(4-S9S749745-*8=*3=8.*=):=:)8 89`=41=1*8=+47U98=8439=)42.3S*8=5&7=)*8=*85T(*8=)&7'7*8= *(942>(47-.?.*33*8= 9*11*= 6:*= )*8= &,&(S*8= *9= )*8= .59*74(&75&(S*8_= 4:8= &;438= (-4.8.= )S9:).*7= )*8= 2>(4-S9S749745-*8= 574(-*8= )*= 2>(4-S9S749745-*8= 9*25S7S*8= 57S(S)*22*39= S9:).S*8`=&+.3=)*=7*89*7=)&38=1*=2U2*=(&)7*=5->14,S3S9.6:*=*9=)*=(425&7*7=1*8=(-&25.,3438= &884(.S8_= 1=8&,.9=) 5->1147(-.8=2439&3&`= _=(&:)&9& =*9=)*= *5-&1&39-*7&=*=.,:&`= 974.8=*85T(*8=)*= *499.S*8=):=8:)8*89=&8.&9.6:*_= 4:9*8= 974.8= 8*= 8439= 7S;S1S*8= &884(.S*8= A= )*8= (-&25.,3438= *(942>(47-.?.*38`= 2&.8= 54:7= 1*8= )*:== 5->1147(-.8 `= )*= +&O43= 343= 85S(.+.6:*_= *8= 7S8:19&98= ;439= A= 1*3(4397*= )*8= 9-S47.*8=&;&3(S*8=8:7=1S;41:9.43=)*=1&=2>(4-S9S749745-.*=*9=548*=)*=342'7*:8*8=6:*89.438= 6:&39= &:= +43(9.433*2*39= )*= (*99*= &884(.&9.43= *9= A= 1&= 8S1*(9.43= )*= (*99*= 343= 85S(.+.(.9S_= &:97*8=2>(4-S9S749745-*8=439=)*5:.8=S9S=S9:).S*8=)&38=1*8=2U2*8=7S,.438=*9=8439=&884(.S*8= 85S(.+.6:*2*39= A= )*8= (-&25.,3438= *(942>(47-.?.*38_= 1= 5*:9= 8&,.7= ):3*= 5&79.(:1&7.9S= ):= ,*37*= 5->1147(-.8 `= :3= ,*37*= *39.T7*2*39= 2>4(-S9S749745-*= (42548S= )*= --= *85T(*8= &8.&9.6:*8`=84.9=:3=)*8=,*37*8=)*=2>(4-S9S749745-*8=1*8=51:8=).;*78.+.S_= 7&7= &.11*:78`= (*9= &79.(1*= )S24397*= 6:*= (*79&.3*8= 47(-.)S*8= 9*77*897*8= 9745.(&1*8= (-14745->11.*33*8= 5*:;*39= 7*(*;4.7= :3*= 5&79= )*= 1*:7= (&7'43*= )*8= (-&25.,3438= *(942>(47-.?.*38= 6:*11*8= 439= *3= (422:3= &;*(= 1*8= &7'7*8= *3;.7433&398_= *8= (&8= )*= 2.=49745-.*=S9&.*39=/:86:&1478=(433:8=:3.6:*2*39=(-*?=)*8=47(-.)S*8=9*77*897*8=9*25S7S*8_= BMC Biology Bio Med Central

Research article Open Access Two mycoheterotrophic orchids from Thailand tropical dipterocarpacean forests associate with a broad diversity of ectomycorrhizal fungi Mélanie Roy* 1, Santi Watthana 2, Anna Stier 1, Franck Richard 1, Suyanee Vessabutr 2 and Marc-André Selosse 1

Address: 1Centre d'Ecologie Fonctionnelle et Evolutive (CNRS, UMR 5175), Equipe Interactions Biotiques, Montpellier, France and 2Queen Sirikit Botanic Garden, Mae Rim, Chiang Mai, Thailand Email: Mélanie Roy* - [email protected]; Santi Watthana - [email protected]; Anna Stier - [email protected]; Franck Richard - [email protected]; Suyanee Vessabutr - [email protected]; Marc-André Selosse - [email protected] * Corresponding author

Published: 14 August 2009 Received: 4 March 2009 Accepted: 14 August 2009 BMC Biology 2009, 7:51 doi:10.1186/1741-7007-7-51 This article is available from: http://www.biomedcentral.com/1741-7007/7/51 © 2009 Roy et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0 ), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract Background: Mycoheterotrophic plants are considered to associate very specifically with fungi. Mycoheterotrophic orchids are mostly associated with ectomycorrhizal fungi in temperate regions, or with saprobes or parasites in tropical regions. Although most mycoheterotrophic orchids occur in the tropics, few studies have been devoted to them, and the main conclusions about their specificity have hitherto been drawn from their association with ectomycorrhizal fungi in temperate regions. Results: We investigated three Asiatic Neottieae species from ectomycorrhizal forests in Thailand. We found that all were associated with ectomycorrhizal fungi, such as Thelephoraceae, Russulaceae and Sebacinales. Based on 13 C enrichment of their biomass, they probably received their organic carbon from these fungi, as do mycoheterotrophic Neottieae from temperate regions. Moreover, 13 C enrichment suggested that some nearby green orchids received part of their carbon from fungi too. Nevertheless, two of the three orchids presented a unique feature for mycoheterotrophic plants: they were not specifically associated with a narrow clade of fungi. Some orchid individuals were even associated with up to nine different fungi. Conclusion: Our results demonstrate that some green and mycoheterotrophic orchids in tropical regions can receive carbon from ectomycorrhizal fungi, and thus from trees. Our results reveal the absence of specificity in two mycoheterotrophic orchid-fungus associations in tropical regions, in contrast to most previous studies of mycoheterotrophic plants, which have been mainly focused on temperate orchids.

Background parasitic taxa, a strategy called mycoheterotrophy (MH) During the last decade, important advances have been has been shown in more than 400 species within several made in our understanding of nutrition of achlorophyl- plant clades, showing patterns of convergent evolution to lous, heterotrophic plants [1]. Beyond the classical plant- heterotrophy [2]. MH plants receive carbon from soil

Page 1 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

fungi colonising their roots, forming the so-called mycor- 90% occur in tropical regions, including a diversity hot rhizal symbiosis [3]. MH in adult (above-ground) life spot in tropical Asia where 120 species grow [4,28]. There phases has evolved repetitively among orchids [4]. Recent have been recent investigations on mycorrhizae from advances were made in identifying mycorrhizal fungi of tropical orchids, but they exclusively focused on green, MH plants by molecular methods, thus revealing their epiphytic species (see, for example, [29,30]). They ultimate carbon source, the photosynthesised carbon of revealed more or less specific associations with the fungal autotrophic plants associated with the same mycorrhizal clades found in autotrophic temperate orchids, the fungi in most cases. 'rhizoctonias' [31]. This group of otherwise parasitic and saprobic fungi encompasses Ceratobasidiales, Tulasnella- PCR amplification and sequencing of the fungal ribos- les and Sebacinales (from clade B sensu Weiss et al. [32]), omal DNA from mycorrhizae allowed identification of and is absent from MH species MH mycorrhizal fungi in more than a dozen MH orchid species [5-9], as well as in several MH species among Eri- Our aim was to investigate the identity of the fungal part- caceae [10], Gentianaceae and Corsiaceae [11], and Bur- ners and specificity of the association of tropical Asiatic manniaceae [12]. All these studies identified a very MH orchids, and to compare the putative origin of their specific association, that is, of each MH species with fungi carbon with that of temperate MH orchids. In this study, from a single genus or even a sub-clade within a genus. two important factors were taken into account. First, ECM Most fungi involved are mycorrhizal partners on other fungi are absent from some tropical forests [3]. We thus autotrophic plants, forming arbuscular mycorrhizae (AM) focused on tropical Asiatic forests that are dominated by [11-13]. As exceptions, some tropical orchids associate ECM Fagaceae and Dipterocarpaceae tree species [3,33]. with saprobic fungi [14-17], but are often specific too. Here, as in temperate forests, AM, ECM, and various Aside from these tropical exceptions, the fungal associates saprobic fungi are available, as well as rhizoctonias asso- of most MH plants suggest that a carbon flow from sur- ciated with green orchids [34]. Second, we focused on MH rounding autotrophic plants to the MH plants, via the species from a clade already studied in temperate regions shared mycorrhizal fungus, is likely to occur. to control for differences resulting from the orchids' phylogenetic position. The Neottieae, in which MH species For temperate MH plants, the stable isotope composition arose several times [35,36], are well studied in temperate of MH plants supports nutrition on ectomycorrhizal regions, where they reveal specific associations with ECM (ECM) fungi. Natural abundances in 13 C and 15 N are fungal clades: Thelephoraceae in Cephalanthera austinae major tools in ecology to detect the food source of an [37], Russulaceae in Limodorum abortivum [38], and Sebac- organism [18,19]. Most organisms have a 13 C abundance inales in Neottia nidus-avis [25,39]. similar to their food source, and indeed MH plants have similar or slightly higher 13 C abundances than associated The tropical Asiatic Neottieae tribe encompasses 33 MH fungi [20,21]. As an exception, however, ECM fungi are species from the enigmatic genus Aphyllorchis [35], and richer in 13 C than autotrophic plants [22]. Although the thus represents one of the most diversified MH genera. reasons for this fractionation are unclear [23], it entails a The position of Aphyllorchis among the Neottieae is still difference in 13 C abundance between autotrophic and not supported by molecular data [35], and even its mono- MH plants [24]. 15 N accumulates along food chains, due phyly is questioned [40]. In this study, we focused on to a fractionation at each trophic level [19], and its abun- three MH species occurring in ECM forests from Thailand, dance usually increases in the order autotrophic plants < namely Aphyllorchis montana , A. caudata and Cephalanthera ECM fungi MH plants [20,24]. Moreover, 14 C labelling exigua (Figure 1). Assuming phylogenetic conservatism for experiments have provided direct evidence that MH the traits under study, and based on temperate species orchid and Ericaceae receive assimilates from surrounding already investigated, we expected them to be specifically trees through shared mycorrhizal fungi [25,26]. associated with narrow ECM clades, and to use tree photosynthates by way of shared ECM fungi. Our aims were, Current investigations are strongly biased toward MH within Thailand forests and for these three species, to test plants from temperate regions. For example, with the these predictions, that is, (i) to confirm that Aphyllorchis exception of a recent study [16,17], few N and C isotopic belongs to Neottieae; (ii) to identify fungal associates of analyses have been performed on tropical MH plants. The the three species; (iii) to infer their fungal specificity level; locations of the laboratories involved, and perhaps the and (iv) to investigate their isotopic content in 13 C and Convention on International Trade in Endangered Species 15 N, to infer their carbon source. of Wild Fauna and Flora [27], may have limited research on MH species in tropical regions. However, dense cover Results in tropical forests, which select for light-independent Phylogenetic position of Aphyllorchis spp. and C. exigua nutrition, provides a useful opportunity to study MH Based on three markers (ITS, trnS-G and rbcL ; GB acces- plants. Indeed, among the ca . 200 MH orchids, more than sion numbers FJ454868 –FJ454884 , Additional file 1), the

Page 2 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

a b c sample, except Resinicium sp. and Malassezia sp. (two samples each). Saprobes were mainly basidiomycetes (69%) and ascomycetes (23%). In all, 15% of orchid individuals did not reveal any ECM fungus, 45% of the individuals revealed a single ECM sequence (sometimes in addition to saprobes and endophytes), and all remaining individu- d e als associated with two to four ECM fungi (Figure 4), sometimes on the same root (Figure 5). Thus, orchid indi- rh viduals had diverse partners (up to nine putative ECM r species from five different genera in a single individual – d AMD7.1, Additional file 2, Figure 5). In seven samples, rh r two different sequences were detected by cloning. In five r 5ȱcm 1ȱcm individuals (12.5%) only, identical fungal sequences were retrieved from different roots.

TheFigure three 1 mycoheterotrophic orchid species Nine A. caudata individuals (from two populations, that The three mycoheterotrophic orchid species . The is, 27 samples; Table 1 and Additional file 2) produced 23 three mycoheterotrophic orchid species under study, A. mon- simple PCR products that belonged to 12 RFLP types, nine tana (a) , A. caudata (b) , and C. exigua (c) , with closer views of which were successfully sequenced (Additional file 2). of the underground parts of A. montana roots (d) and C. exi- Four multiple PCR products were cloned and produced gua (e) . Abbreviations: r, root; rh, rhizome. eight different sequences (Additional file 2). Apart from an endophyte, found only once (Hypocrea sp.; Figure 3b and 3d), all sequences were putatively ECM, and mainly Neottieae tribe was monophyletic and included the two belonged to Russulaceae, Thelephoraceae and Sebacinales Aphyllorchis under study (Figure 2). Thaia saprophytica , a (from the ECM-forming clade A, sensu Weiss et al. [32]). green species from Thailand, had a basal position, but two Among all individuals, 55% displayed a single ECM fun- markers ( rbcL and trnS-G ) were not obtained for this spe- gus, whereas 45% displayed two to three ECM fungi. As cies and this limited the support level. Identical topolo- for A. montana, no fungal taxon was shared by all individ- gies at genus level were found, although with lower uals. support levels, when using the three markers separately (data not shown). The genera Epipactis , Listera and Nine C. exigua individuals (72 samples) from one popu- Cephalanthera were monophyletic, but this, together with lation produced 63 simple PCR products belonging to 16 their relative positions, remained weakly supported. The RFLP types that were all sequenced. Putative ECM fungi position of C. exigua within the genus Cephalanthera was dominated the fungal community (84%), with some well supported, and the two Aphyllorchis species clustered rhizoctonias (5%) and saprobes (11%; Figure 3c and 3d). together as a well-supported sister clade to the European Thelephoraceae represented 65% of identified fungi, and genus Limodorum . one (FJ454907 ) was even found in 16 samples arising from seven individuals. Putative ECM Helotiales and Nau- Molecular identification of root fungi coria sp. were found in one sample each, as well as Forty A. montana individuals (from seven populations, saprobes including Trichoderma sp. (8% of all fungi) and that is, 288 root samples; Table 1[41] and Additional file other ascomycetes (in one sample each). In all, six out of 2) produced 220 simple PCR products, representing 135 nine individuals exclusively associated with Thelephora- restriction fragment length polymorphism (RFLP) types, ceae, two displayed two different ECM fungi, with a dom- 104 of which were successfully sequenced. In addition, we inance of Thelephoraceae (+80% of the samples), and one successfully cloned eight multiple PCR products that pro- displayed only Helotiales. Thus, Thelephoraceae were the duced 11 different sequences (Additional file 2). BLAST preferred fungal associates of C. exigua . identifications showed that 83% were putatively from ECM fungi, 4% from rhizoctonias, 3% from endophytes Molecular identification of A. montana fungal pelotons and 10% from saprobic fungi (Figure 3a and 3d; Addi- The identity of fungi colonising mycorrhizal cells was tional file 2). ECM fungi belonged to diverse taxa, mainly assessed on peloton pools (pools of twelve pelotons from Russulaceae, Thelephoraceae and Clavulinaceae. Endo- a single root section) from two A. montana individuals at phytic fungi and Thanatephorus sp., a typical orchid myc- Doi Suthep #2 (Table 1 and Additional file 2). On orrhizal fungus, were found each from a single sample, on AMD6.1, two pools revealed a Helotiales (FJ454973 ) individuals also displaying ECM fungi. Thirteen different already found on the same individual, and four revealed a saprobes were identified, each occurring only on a single Russulaceae (FJ454956 ) already found on other Doi

Page 3 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

same root (Figure 5). Considerable internal transcribed spacer (ITS) variations in Thelephoraceae (also very fre- Epipactis tremolsii tremolsii quent, 30.0% of the sequence found on Aphyllorchis spp. Epipactis lusitanica lusitanica

Epipactis leptochila leptochila and 63.2% on C. exigua ) forbade phylogenetic analysis,

Epipactis fibrii fibrii but sequences were not more similar within than between Epipactis duriensis duriensis orchid species (data not shown). 74 Epipactis muelleri muelleri

Epipactis flava flava

Epipactis helleborine helleborine Using a threshold of < 97% of ITS variation to delineate

Epipactis microphylla microphylla species, 112 species were recorded in A. montana , 23 in A.

Epipactis albiensis albensis 70 caudata and 31 in C. exigua . In all, 94% of these species Epipactis fageticola fageticola were represented by a single sequence. Only four species Epipactis palustris palustris

100 AphyllorchisAphylloçrchis caudata caudata occurred on more than one individual, and were all from 82 Aphyllorchis montana montana the same A. montana population (Additional file 2). Rare-

Limodorum abortivum abortivum 72 faction analyses provide similar trends when (i) consider- Cephalanthera longifolia longifolia 86 ing either all fungi or ECM fungi only; (ii) making the Cephalanthera damasonium damasonium 100 Cephalanthera exigua exigua analysis at fungal family or species level; and (iii) pooling Cephalanthera rubra rubra all populations or separating them to calculate mean val-

NeottiaListera ovata ovata 90 ues for each species. In every case, curves for A. montana Neottia nidus-avis nidus-avis 72 and A. caudata were similar (Figure 7a), and higher than NeottiaListera smallii smallii

Palmorchis trilobulata for C. exigua , so that the low fungal diversity in this species

Thaia saprophytica saprophytica 94 was not a sampling artefact. In detrended component Tropidia polystachya polystachya analysis (DCA), no differences in ECM fungal community Nervilia shinensis shinensis were found between A. montana populations (data not Vanilla planifolia planifolia

0.03 0.03 substitution / site shown) or between the two Aphyllorchis species (Figure 7b). In contrast, the C. exigua ECM fungal community differed from two Aphyllorchis species ( P < 0.01 for both tests; PhylogeneticFigure 2 tree of the Neottieae tribe Figure 7b). Neither the forest type nor the geographical Phylogenetic tree of the Neottieae tribe . Phylogenetic origin had a significant effect ( P > 0.05; data not shown). tree of the Neottieae tribe showing positions of A. montana , Results were unchanged when considering all fungi. Thus, A. caudata and C. exigua ; Mycoheterotroph species are in C. exigua strongly differed in fungal community structure bold. Phylogeny based on a concatenation of ITS, trnS-G and rbcL , using the maximum likelihood method (general time from the two Aphyllorchis species, both quantitatively and reversible model). Numbers on branches indicate bootstrap qualitatively. values above 70% (over 1,000 replicates). Stable isotope analyses We tested by analyses of natural content in stable isotopes Suthep #2 individuals (Additional file 2). On AMD7.1, and C/N ratio whether ECM fungi were potential C two fungi already found on the same individual were sources for the MH orchids. At Doi Suthep #2 (Figure 8a), recovered, namely a Clavulinaceae (FJ454977 ; three significant differences for both !13 C and !15 N occurred in pools) and a Thelephoracae (FJ454979 ; one pool), while the order autotrophic Boesenbergia rotunda < other cloning on another pool revealed a mix of the two previ- autotrophic plants < A. montana ECM fungi (including ous fungi and a Russulaceae (FJ623066 , close to R. illota taxa found on A. montana roots, Russulaceae and Thele- and some Russulaceae already found at Doi Suthep #2, phoraceae). C/N ratio values were higher for autotrophs Additional file 2). On both individuals, four pools did not than for fungi (12.1 ± 1.2 – mean ± SD) and A. montana amplify. These data corroborated that (i) several ECM (11.9 ± 1.2; Figure 9a): the latter two were not signifi- fungi were mycorrhizal on the same individual, even the cantly different (Mann-Whitney test, P = 0.81), but signif- same root, and (ii) ECM asco- and basidiomycetes were icantly lower than autotrophs (22.6 ± 3.0 on average, P < mycorrhizal on A. montana . 0.001; B. rotunda did not differ from other autotrophs in this respect; Figure 9a). !13 C values and variations in !15 N Analysis of the fungal community analysis and C/N ratio were congruent with a food chain from Russulaceae, by far the most represented on A. montana autotrophs to ECM fungi and A. montana . and A. caudata (39.8% of the sequence found, in 33.6% of typed samples), were phylogenetically over-dispersed At Doi Suthep #3, !13 C was higher for A. caudata than for (Figure 6), further supporting the low specificity of myc- autotrophic plants but (not significantly) lower than for orrhizal association. Even fungi identified from the same saprobic fungi. A. caudata had !15 N intermediate between individual did not cluster together (data not shown), and the different saprobic fungal species, but higher than different Russulaceae species sometimes colonised the autotrophs (among them, the orchid Ludisia discolor had

Page 4 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

Table 1: Description and location of sampling sites.

Species and sampling site Geocodes and elevation Type of forest and dominant trees " No. of orchids sampled

Aphyllorchis montana Doi Suthep #1 18°48'39" N Evergreen forest: Castanopsis acuminatissima (F) , 11 98°55'00" E Castanopsis diversifolia (F) , Dipterocarpus costatus (D) , 1053 m Manglietia garretti (Magnoliaceae) , Carallia brachiata (Rhizophoraceae). Doi Suthep #2 18°48'24" N Evergreen forest: Dipterocarpus costatus (D) , Castanopsis 11 98°55'19" E diversifolia (F). (+ isotopes samples) # 950 m Queen Sirikit Botanical Garden #1 18°54'24" N Dry dipterocarpacean forest: Dipterocarpus costatus (D) , 8 98°51'48" E Shorea roxburghii (D) , Castanopsis argyrophylla (F) , 811 m Castanopsis tribuloides (F) , Lithocarpus thomsonii (F), Diospyros variegate (Ebenaceae) , Phoebe lanceolata (Lauraceae) , Protium serratum (Burseraceae). Queen Sirikit Botanical Garden #2 18°53'36" N Oak forest: Lithocarpus sootepensis (F) , Dipterocarpus 2 98°51'27" E costatus (D) , Shorea roxburghii (D). 729 m Nam Nao 16°93'48" N Bamboo forest: Castanopsis diversifolia (F) , Lithocarpus sp 1 101°33'39" E (F). 700 m Khao Chamao 12°58'41" N Dipterocarpacean forest: Dipterocarpus dyeri (D). 4 101°42'05" E 800 m Klong Pla Kaeng 12°56'08" N Dipterocarpacean forest: Dipterocarpus dyeri (D). 2 101°44'09" E 700 m

Aphyllorchis caudata Doi Suthep #3 18°48'39" N Evergreen forest: Dipterocarpus costatus (D) , Castanopsis 11 98°55'00" E diversifolia (F). (+ isotopes samples) # 1050 m Doi Inthanon 18°35'25" N Evergreen forest: Castanopsis acuminatissima (F), 2 98°29'09" E Lithocarpus sp. (F), Quercus sp. (F) , Schima wallichii 1000 m (Theaceae).

Cephalanthera exigua Doi Pee Pan Nam 19°06'05" N Evergreen forest: Castanopsis acuminatissima (F) , 9 99°20°84"E Castanopsis sp. (F) , Gironniera sp. (Ulmaceae), Lithocarpus (+ isotopes samples) # 2015 m sp. (F) , Michelia floribunda (Magnoliaceae), Myrica esculenta (Myricaceae), Neolitsea sp. (Lauraceae), Camellia oleifera (Theaceae) , Schima wallishii (Theaceae), Syzygium angkae, Syzygium sp. (Myrtaceae).

" F: Fagaceae; D: Dipterocarpaceae. # For species sampled for isotopic studies, see Figure 6. significantly higher !15 N). Unfortunately, no ECM fungi ECM fungi plus the orchids Cheirostylis montana

Page 5 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

a c Saprobes Saprobes Thelephoraceae Endophytes 11 % 11 % 65 % 3 % Rhizoctonia Rhizoctonia Russulaceae 5 % 4 % 37% Other ECM <1 % Naucoria sp. 2 % 3 % Cortinariaceae 1 % Amanitaceae Helotiales 1 % 16 % Heliotaceae 2 % Coltriciellaceae 2 % Sebacinales 2 % Clavulinaceae 13 % Thelephoraceae d 100% 24 %

b Other ECM <1% Endophytes 4 % 4 % 80% Clavulinaceae 4 % Sebacinales 60% 9 %

40% Thelephoraceae 13 % 20% % of fungi identified per species Russulaceae 66 % 0% AM AC CE

DescriptionFigure 3 of the community of fungi identified in the three mycoheterotrophic species roots Description of the community of fungi identified in the three mycoheterotrophic species roots . Frequency of occurrence of fungal taxa identified in each investigated orchid species grouped on a family/order basis for A. montana (a) , A. caudata (b) , and C. exigua (c) , or grouped by ecology ( (d) AM: A. montana ; AC: A. caudata ; CE: C. exigua ; ectomycorrhizal taxa are represented by black lines, rhizoctonias by white dots on black background, endophytes by black dots on white background, and saprophytes in white).

Discussion high specificity hitherto found in all investigated MH spe- We show for the first time that (i) at least some Aphyllor- cies [1], and especially MH orchids [9], mycorrhizal asso- chis belong to the Neottieae tribe; (ii) tropical (Thailand) ciations in the two Aphyllorchis species studied here Neottieae associate with ECM fungi; and (iii) that they are revealed a very low specificity, while C. exigua proved to likely to use their ECM fungi (and thus nearby trees) as a be more specific. C source. This is congruent with what is known from temperate Neottieae species [5,20,21,37], but we provide here Tropical MH Neottieae associate with ECM fungi the first isotopic evidence that tropical MH orchids associ- In temperate regions, ECM fungi consistently associate ate with ECM fungi. Furthermore, in sharp contrast to the with roots of Neottieae, both green [36,41-44] and MH

Page 6 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

100% dominating on C. exigua . Russulaceae are specific associ- Thelephoraceae sp. #35 Thelephoraceae sp. #7 ates of the Mediterranean Limodorum abortivum [38], a sis- Thelephoraceae sp. #33 Thelephoraceae sp. #32 ter species to the genus Aphyllorchis (Figure 2); however, Thelephoraceae sp. #13 80% Thelephoraceae sp. #12 the species found here were unrelated to the R. delica clade Sebacina sp. #5 Sebacina sp. #3 mycorrhizal on L. abortivum (Figure 7). C. exigua specifi- Russulaceae sp. #46 Russulaceae sp. #37 cally associated with Thelephoraceae, which are specific Russulaceae sp. #8 C. austiniae 60% Russulaceae sp. #35 associates of the related North American MH Russulaceae sp. #25 and colonise, although not exclusively, green European Russulaceae sp. #35 Russulaceae sp. #11 and Asiatic Cephalanthera spp. [38,41,42,44,50]. The exist- Russulaceae sp. #10 Russulaceae sp. #9 40% ence of some phylogenetic inertia in fungal preference Russulaceae sp. #13 Cortinarius sp. #1 within Neottieae (or even within the genus Cephalanthera ) Clavulina sp. #9 Clavulina sp. #8 is an appealing possibility that deserves further study, Clavulina sp.#7 Clavulina sp. #5 20% including more species and a more robust phylogeny of Clavulina sp. #4 Clavulina sp.#3 Abundance (%) in each orchid individual this tribe. With the possible reversion of some Epipactis sp. Clavulina sp. #2 Athelia sp. #1 [36,42], we confirm here that the Neottieae lost associa-

0% tion with the rhizoctonias (the plesiomorphic mycor- AMQ1 AMQ2 AMQ3 AMQ4 AMQ5 AMQ6 AMQ9 AMQ11 rhizal feature among orchids) and became associated with alsDifferencesFigure from 4 a given in ectomycorrhizal population taxa between orchid individu- ECM fungi irrespective of their global localisation (a Differences in ectomycorrhizal taxa between orchid Cephalanthera longifolia individual from a Myanmar forest individuals from a given population . Diversity and abun- also revealed ECM fungi, including Russulaceae – GB dance of ectomycorrhizal taxa identified in 10 individuals accession numbers FJ454917 –FJ454919 , see Figure 6). (AMQ1 to 10) from an A. montana population (Queen Sirikit Botanical Garden #1). Tropical MH orchids offer considerable diversity in ecology of associated fungi. ECM fungi have already been found in some tropical MH orchids, such as Lyophyllum [5,25,37,39]. Here, we found that most root fungi had a shimeji (in Erythrorchis ochobiensis [51]) or ECM Ceratoba- putative ECM ecology. Although endophytes, saprobes sidiaceae (in Chamaegastrodia sikokiana [52]). Most species and some rhizoctonias were also found, peloton analysis associate with non-ECM fungi, that is, parasites [53] or in A. montana only recovered ECM fungi. The mycorrhizal saprobes [14-17,54,55], a fungal ecology never found in status of putative endophytes and saprobes remains ques- temperate MH orchids. This fungal diversity is reflected in tionable, as in previous studies on Cephalanthera spp. the fact the MH Gastrodia nana and Epipogium roseum , both [36,41]. ECM fungi are common in Dipterocarpaceae mycorrhizal with saprobic fungi [14,16,17], also occur in and/or Fagaceae forests of South-East Asia [45,46], and the Thailand forests where this study was carried out especially of Thailand [47]. The most frequent taxa in this (Watthana and Roy, personal observations). In this frame- study (Russulaceae, Thelephoraceae) are also the most work, it is tempting to speculate that other factors, such as abundant under Dipterocarpaceae [45,47], where Cla- contingency or phylogenetic inertia, contribute to the vulinaceae and Sebacinales clade A are also known [48]. ecology of the fungus in tropical orchids. For Neottieae, While temperate Neottieae often associate with taxa form- the previously mentioned shift from rhizoctonias to ECM ing hypogeous fruit bodies (such as Tuber or Hyme- fungi [36,42] allowed diversification in ECM forests, not nogaster ; [49]), little evidence for this trend was found only in temperate regions where such forest dominates, here (with the possible exception of sequence FJ454490 but also in tropical forest harbouring ECM trees. The anal- on C. exigua , closely related to the hypogeous Arcange- ysis of mycorrhizal partners in the few Neottieae occurring liella , Figure 6). Relatively wet conditions in the investi- in tropical America and Africa, as well as in some of the gated forests may explain this, since hypogeous taxa have other 33 Aphyllorchis species in tropical Asia [35], is now been shown to be adapted to dry environments [3]. Yet, pending, to allow the construction of a global phylogeo- hypogeous taxa such as Sclerodermataceae exist in Thai- graphic scenario for the Neottieae. land [47]. Moreover, the absence of hypogeous taxa remains difficult to confirm (i) from sequencing data Tropical MH Neottieae likely receive C from nearby ECM only, and (ii) in a context where the fungal diversity trees remains poorly explored. Since ECM fungi almost exclusively receive C from host trees [3], the investigated MH species may indirectly A. montana and A. caudata harboured a highly diverse exploit the nearby trees, by way of mycelial links. This was ECM community (Additional file 2, Figure 5, Figure 6), described for temperate species [5,39,56], and corrobo- very similar for the two species (Figure 6, Figure 7b), dom- rated by the high, fungal-like 13 C and 15 N in MH plants inated by Russulaceae and Thelephoraceae, the latter also [20,24]. Here, our isotopic analyses show similar patterns,

Page 7 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

a !13 C than MH plants [21]. Whether the differences rhizome observed here are specific to these tropical models or the roots result of ECM sampling unrepresentative of the mycorrhizal species is an open question. However, the difference in !13 C between MH and autotrophic plants (ranging Russulaceae ȱ#36 from 6.8‰ to 9.9‰) was in the range observed in temperate ecosystems (+6.9 ± 1.5‰ [24]), whereas more Thelephoraceae ȱ#12, ȱ#13 diverse values were found for MH orchids associated with Russulaceae ȱ#35 Thelephoraceae #32 saprobic fungi (up to +12‰ [16,17]).

Thelephoraceae ȱ#13 15 Russulaceae ȱ#13 Investigated MH orchids tended to have higher ! N and Russulaceae ȱ#8, equal to lower C/N ratio values than ECM fungi, as Thelephoraceae ȱ#33 1ȱcm expected between two consecutive levels in a food chain, respectively due to isotopic fractionation for 15 N [18,20] and the loss of respiratory C [41]. In all, the isotopic data are congruent with a C flow from autotrophs to MH plants b rhizome by way of shared fungi. Since they do not exclude other roots scenarios, only a direct labelling of tree photosynthates

Gymnomyces would definitively assess whether mycelial links between Amanitaceae #2 trees and orchids allow a flow to MH plants. In this regard, Thelephoraceae ȱ the putative scenario and C and N data obtained here do #10, ȱ#31 not differ from those observed in temperate MH orchids. Russulaceae ȱ#12 The existence and roles of common mycorrhizal networks Clavulinaceae # ȱ6 have often been speculated in tropical ecosystems [58], Thelephoraceae ȱ Clavulinaceae #10 but rigorous demonstration is still lacking: inter-plant C #9 Russulaceae ȱ#47 transfers are striking indirect evidence of their existence Thelephoraceae ȱ#9 Clavulinaceae #10 [26]. 1ȱcm

Mixotrophy in tropical orchids In temperate regions, green plants phylogenetically DistributionFiguretana 5 of identified fungi on the root system of A. mon- related to MH plants recover part of their C from their Distribution of identified fungi on the root system of mycorrhizal fungi, especially among orchids A. montana . Diagram of fungal colonisation on two A. mon- [21,38,42,46]. This photosynthetic and partially MH tana root systems, on Q1 (a) and on D7.1 (b) . Numbers nutrition, also called mixotrophy, is considered as an correspond to putative species identified (see Additional file adaptation to understorey conditions, with low light lev- 2). Slashed areas display two different fungi, identified on the els. It can thus be expected in dense tropical forests, but same 1 to 2 mm-thick root section. has not yet been demonstrated [2]. Mixotrophy entails 13 C and 15 N natural abundances intermediate between those of fully autotrophic and MH plants [21,41,59]. congruent with C transfer from trees to MH species, via Here, Cheirostylis montana at Doi Pee Pan Nam had 13 C ECM fungi, for tropical sites. abundance significantly differing from autotrophs and closer to that in ECM fungi and A. caudata . Since the 13 C As in temperate ecosystems, !13 C were higher for fungi content (-27.1 ± 1.5‰) is too low for a C4 photosynthetic than for autotrophs [57]; unfortunately, the sampling did metabolism [18], mixotrophy is likely to occur. A linear not allow comparison between saprobic and ECM fungi two-sources mixing model [60], with mean !13 C values of on each site. Values of !13 C tended to be equal or higher autotrophs and MH plants as references, suggests that for MH orchids as compared with ECM fungi at Doi 82% of its C was of fungal origin (significantly different Suthep #2 (-25.1‰ vs -27.6‰) and Doi Pee Pan Nam (- from zero based on 95% confidence intervals). 23.1‰ vs -26.2‰). At Doi Suthep #3, where no ECM fungi were available, saprobic fungi were higher in !13 C. Mycorrhizal partners of Cheirostylis montana have not been Since saprobes usually tend to have higher !13 C than ECM investigated yet, but deserve further attention. Indeed, fungi [22,57], this site may not contradict the common most research on tropical orchid mycorrhizae deals with trend at the two others. Although it is often assumed that epiphytic species, and only a few terrestrial species have !13 C are identical in ECM fungi and MH plants [20], some been studied, using in vitro isolation techniques that ECM fungi from the same site can be 1 to 2‰ lower in revealed only rhizoctonia fungi [61]. However, several

Page 8 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

ECM fungi are difficult or impossible to isolate [6], and orchids are specific (for example, [14,52]). Thus, tropical therefore molecular approaches are strongly recom- MH orchids exhibit different specificity levels, as reported mended in future attempts to identify fungi of tropical ter- for tropical green epiphytic orchids [66,67]. restrial orchids. Using such approaches, tropical ecosystems may provide model systems for the examina- There are two caveats to the conclusion of non-specificity. tion of mixotrophy in diverse species beyond orchids. Firstly, we do not know whether all or only some of these fungi are providing C: functional specificity cannot be Absence of mycorrhizal specificity in Aphyllorchis spp ruled out. Nevertheless, no constant partner was identi- The lack of mycorrhizal specificity in A. montana and A. fied, suggesting that several different fungi can provide C. caudata is unexpected for MH species. Such a low specifi- Secondly, MH nutrition is also taking place at germina- city, observed both at population and individual levels tion and early seedling development in orchids, since (Figure 4), is very unusual among orchids [31], but has seeds have very few reserves: we do not know whether been found in some mixotrophic Neottieae in the genera seedlings exhibit fungal specificity. In Cephalanthera spp., Cephalanthera and Epipactis [36,41-44]. Aphyllorchis spe- only a subset of fungi present in adult plants are efficient cies associated with various ECM fungi at the population, at this stage [44], and some orchids change or diversify individual, root and cell levels (Figure 5), and no obligate their partners over their lifespan [6]. Indeed, if seedlings or constant partner was identified. In contrast, C. exigua also have a large host spectrum, Aphyllorchis spp. may not associated quite specifically with Thelephoraceae, and be limited by availability of fungal partners. They are suggested that our design did allow detection of specifi- widespread but remain rare, with loose populations (indi- city. Moreover, rarefaction curves confirmed that for what- viduals are often separated by a few meters [28], Roy and ever sampling effort, C. exigua presented a lower diversity Watthana, personal observations). Thus, a different spe- (Figure 7a). Since Aphyllorchis fungal communities look cificity in early life stages cannot be excluded in Aphyllor- like ECM communities from tropical regions [46,62], they chis , and requires further investigation. Several may even reflect a random sampling of available ECM observations of the association during in situ germination fungi by orchids' roots. However, given our limited were obtained in temperate regions, after sowing seeds in knowledge of ECM diversity in Thailand forests, we do not mesh bags [31], but this remains to be applied in tropical know whether there is over- or underrepresentation of ecosystems. ECM fungal taxa colonising nearby trees. Interestingly, Sclerodermataceae, which are common in Thailand dipte- Why is fungal specificity low in tropical MH orchids? rocarpacean forests [47], were absent in orchid roots. We Interestingly, the few non-specific MH plants reported so thus cannot exclude some limited mycorrhizal preference far occur all in tropical ecosystems [17,64,65]. Although in Aphyllorchis spp. An intriguing consideration is that spe- this may be pure coincidence, it may suggest some partic- cificity in green or MH orchids, like C. exigua (Figure 1), ular features of MH plants and/or fungal communities in correlates with short roots (less than 10 cm in length, or tropical ecosystems. Specificity in biological interactions even absent; Roy, personal observation), whereas non- reveals variable latitudinal patterns, ranging from higher specific species, such as Aphyllorchis spp., have long roots specificity in the tropics (for example, for plant endo- (up to 50 cm; Figure 1). phytic fungi [68]) to similar or lower specificity (for example, for phytophageous and pollinating insects [69]). Aphyllorchis species contrast with the highly specific tem- Difference between latitudes thus relates more to the func- perate MH Neottieae studied so far, such as Neottia nidus- tioning of each interaction. However, the raison d'être of avis (with Sebacinales [5]) and C. austinae (with Thele- MH specificity remains poorly understood in temperate phoraceae [37]). Ironically, Aphyllorchis' closest phyloge- MH species. Two non-excluding models were proposed, netic relative, the mixotrophic Limodorum abortivum namely functional co-adaptation and parasitic co-evolu- (Figure 2), specifically associates with the Russulaceae tion [70]. Functional co-adaptation states that the mecha- [38]. A strong trend toward specificity is reported in nearly nism reversing the C flow (which goes from plant to all MH plants [1,6]: individuals are associated with a nar- fungus in common mycorrhizae [71]) requires fine plant row fungal clade of fungus, and specificity results in local adaptations to fungal physiology, and that specific adap- specialisation or even specialisation toward distinct geno- tations are better than universal ones (functioning with types within populations [7,63]. To the best of our knowl- any fungus). However, the many shifts of fungal partners edge, the only reported exceptions are (i) two taxa of AM during the evolution of MH lineages [8,10] are not pre- fungi in MH African Burmannia congesta and Sciaphila led- dicted by this model. Parasitic co-evolution assumes that ermannii [64]; (ii) saprobic Basidiomycetes in the Carib- MH plants parasitise their mycorrhizal fungus (and thus bean MH Wullschlaegelia aphylla [17], an orchid distantly 'epiparasitise' on green plants [1]), although there is no related to Neottieae; and (iii) the case of another MH direct evidence of detrimental effects [39,70]. In this case, orchid, Erythrorchis cassythoides [65]. Other tropical MH specificity would evolve within an arms race between the

Page 9 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

FJ455016FJ455016 AM FJ455015 AM 100100 FJ455015 AM FJ454987 AM 59 FJ454987 AM FJ455017FJ455017 AM 97 9999 97 FJ455010FJ455010 AM 78 78 LactariusLactarius corrugis corrugis EU598154 LactariusLactarius volemus volemus AY606959 100100 FJ455020FJ455020 AM 9686 100 FJ455021FJ455021 AM 80 84 80 FJ455022FJ455022 AM LactariusLactarius gerardii gerardii EF560688 100100 FJ454998FJ454998 AC RussulaRussula vesca DQ422018 9797 Russula heterophylla AY061681 9494 Russula heterophylla FJ454925FJ454925 AM RussulaRussula parazurea DQ422007 9696 RussulaRussula aeruginaaeruginea DQ421999 RussulaRussula pallescens DQ421987 RussulaRussula cyanoxantha DQ422033 DQ422033 100100 RussulaRussula cyanoxantha DQ974758 DQ9747580 RussulaRussula favrei EF530944 100100 FJ454917FJ454917 CL FJ455026FJ455026 AM RussulaRussula brevipes AF349714 LactariusLactarius longisporus longisporus DQ421971 RussulaRussula earlei DQ422025 RussulaRussula albonigra DQ422029 RussulaRussula subnigricans EF534351 RussulaRussula densifolia AF418606 100100 FJ454902FJ454902 AC 100100 FJ454969FJ454969 AM RussulaRussula nigricans AB154708 AB154708 9797 RussulaRussula nigricans AM087260 AM087260 70 70 FJ454996FJ454996 AM 100100 FJ454990FJ454990 AM FJ454938FJ454938 AM FJ454960FJ454960 AM FJ454958FJ454958 AM FJ454956FJ454956 AM FJ454961FJ454961 AM 100100 FJ454957FJ454657 AM FJ454965FJ454965 AM RussulaRussula illota DQ422024 RussulaRussula laurocerasi AY061735 FJ454982FJ454982 AM 100 RussulaRussula foetens AY061677 FJ454959FJ454959 AM RussulaRussula virescens EU819437 RussulaRussula alboareolata AF345247 FJ454929FJ454929 AM FJ454926FJ454926 AM FJ454924FJ454924 FAM 100100 8080 FJ454927FJ454927 AM 83 Russula variata AB154732 83 Russula variata AB154732 RussulaRussula variata EU819436 EU819436 FJ454988FJ454988 AM FJ454968FJ454968 AM 9898 FJ454967FJ454967 AM 9999 FJ455032FJ455032 AM 7272 RussulaRussula azurea AY061660 RussulaRussula rosea AY061715 9999 RussulaRussula rosacea AF345249 9191 RussulaRuccula flavida EU598171 A. montana

7878 RussulaRussula vinacea EU598181 Doi Suthep #1 8888 FJ454998FJ454698 AC Doi Suthep #2 RussulaRussula sanguinea AY061718 QSBG 84 Russula aquosa AY061657 Russula aquosa Nam Nao 84 Russula raoultii AF418621 Russula raoultii Khao Chamao 100100 RussulaRussula compacta EU598172 LimodorumLimodorum abortivum abortivum isolate fungus DQ061931 DQ0611931 Khao Pla Kaeng 7171 LimodorumLimodorum abortivum abortivum isolate fungus DQ061930 DQ0611932 100100 A. caudata RussulaRussula delica AF418605 Doi Inthanon FJ454490FJ454900 ACAC 9090 ArcangeliellaArcangeliella camphorata camphorata EU644702 ZelleromycesZelleromyces hispanicus hispanicus AJ555567

0.3 0.3 substitution / site

Over-dispersionFigure 6 of Russulaceae isolated from A. montana and A. caudata Over-dispersion of Russulaceae isolated from A. montana and A. caudata . Unrooted phylogenetic tree placing the Russulaceae identified from Aphyllorchis montana (AM) and A. caudata (AC). This phylogeny is based on internal transcribed spacer sequences, using maximum likelihood (general time reversible model). Numbers on nodes indicated bootstrap values above 70% (over 10,000 replicates).

Page 10 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

a orchids. In both cases, the C demand would not be very costly for the fungus. We respectively assume that (i) mix- AM otrophic plants have limited C requirements, because of their photosynthesis, and (ii) due to better tree photosynthesis (higher primary production) in the tropics, tropical ECM fungi receive a greater C flow. In both cases, the C uptake would be relatively negligible, as compared with the C demand of MH plants on temperate ECM fungi. 26 Thus, functional co-adaptation and/or parasitic co-evolu- 19 Number of species tion would not apply in tropical regions since avoidance AC 15 7 CE mechanisms are selected only if the cost of avoidance is 6 0 20 40 60 80 100 80 60 40 20 0 lower than the cost of interaction [72]. This statement 22 30 Number of samples (root sections) investigated remains speculative, since we know little about the C budget in individual mycelia, and comparative fungal b physiology in tropical versus temperate regions. More studies of orchid-fungal diversity in tropical ecosystems are required to support it. Making this assumption, spe- CE cific MH plants and also some specific temperate mixotrophic plants (such as Limodorum abortivum [38]) would AC simply go beyond a threshold in terms of C loss for the AM fungus, thus entering the co-adaptation and/or parasitic Axis 2 co-evolution process leading to specificity. Alternative explanations remain possible: heterogeneous environments make generalists fitter than specialists [73,74]. Axis 1 Unfortunately, we do not know the structure and spatial heterogeneity of ECM at our sampling sites, and there is DifferencesFigurefungal communities 7 between the three mycoheterotrophic orchid even some evidence that tropical ECM communities are Differences between the three mycoheterotrophic less diverse than temperate ones (K Nara, personal com- orchid fungal communities . Comparison of the fungal munication). communities found on the three orchid species (AM, A. montana , AC, A. caudata and CE, C. exigua ). (a) Rarefaction Conclusion curves for ectomycorrhizal fungal species. (b) Detrended All Neottieae examined to date in both temperate and, component analysis of orchid individuals plotted in two dimensions, based on ectomycorrhizal fungal communities now, tropical ecosystems have been found to associate (fungal taxa grouped by families; note that most points are with ECM fungi. In most cases, they receive C from ECM superimposed). White circles: A. montana individuals; grey mycelial networks linking them to nearby trees, as shown circles: A. caudata ; black triangles: C. exigua . Large symbols by their isotopic content. During Neottieae evolution, represent means for each species, with standard deviations. specificity arose repeatedly, but unexpectedly this turns out to be unrelated to full MH nutrition; in spite of several shifts in fungal partners, some phylogenetic inertia may fungus and the MH plant: first, epiparasitic plants can have occurred. The lack of specificity is encountered for a only associate with exploitable fungi that are somehow few other tropical MH plants, suggesting that MH and resistant to epiparasitism (non-resistant fungi may not fungal organisms from tropical ecosystems may differ support epiparasites and the association could not be functionally from their temperate analogues. This and the maintained), then both partners may select for adapta- observation of mixotrophy in green orchids calls for more tions reducing the cost of this association, and such adap- focus on mycorrhizal associations of terrestrial herba- tations makes the association more and more specific. As ceous plants in the tropics, to know more on the taxo- a result, few co-evolved plant-fungus combinations are nomic position of their fungi and functional diversity successful, and evidence for local adaptation in MH pop- (especially in terms of C flow) of their mycorrhizal associ- ulations [10] and co-evolution with fungi [12,63] support ation. this. Our study and a few others [18,64,65] suggest that these mechanisms at least do not apply to tropical MH Methods plants. Model species and sampling sites Aphyllorchis montana Rchb.f., A. caudata Rolfe ex Downie We propose a common reason to explain non-specificity and Cephalanthera exigua Seidenf. are MH orchids (Figure in (i) any mixotrophic plants, and (ii) tropical MH 1) from South-East Asia that grow in low to high moun-

Page 11 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

1212 a Theleph ora sp. (c, e ) 1010 Aphyllorchis montana (d , e ) 8 8 Tomentella sp. (c, d) ‰ 6 6 Lactarius sp. 2 (c,c ) Lactarius sp. 3 (c,c ) Lactarius sp. 1 (c,c ) Cortinarius sp. (c,c )

15N 4 4 δ 2 2 Castan opsis acuminatissima (b,b ) 0 0 Canthium inerme (b,b ) Smilax sp. (b,b ) -2 -2 Bo esenbergia r otunda (a , a) 110-04 b 88 Unkown sapr obe (d , g) Aphyllorchis caudata (d , f) 66

‰ Collybia sp. (d , f) 44 Ludisia disc olor (a , e ) 15N

δ 22 Lasianthus kursii (c, c ) Stereum sp. (d , d) 00 Commelina c ommunis (a , c ) Macaranga k ursii (b, b )

-2-2 Acanthace ae sp. (a , a)

1414-4 c 1122 Cephalanthera exigua (c, f) 1100 Theleph ora sp. (b, e ) 88 Marasmius sp . (b, d) ‰ 66 Laccaria sp . (b, c ) 4

15N 4 δ 22 Cheir ostylis m ontana (b, b ) 00

-2-2 Smilax sp. (a , a) Aspidista elati or (a , a) Impatiens sultani (a , a) -4-4 Str obilenthes dyerianus (a , a)

-6-6 -40 - 38 -36 - 34 -32 - 30 -2 8 -26 -2 4 -22 -20 δ13C ‰

IsotopicFigure 8 signature of the three mycoheterotrophs studied and other green orchids Isotopic signature of the three mycoheterotrophs studied and other green orchids . Carbon versus nitrogen stable isotope values (‰) of green plants, mycoheterotrophic plants (names bold) and fungi (names underlined) at (a) Doi Suthep #2 (including A. montana and various ectomycorrhizal (ECM) fungi), (b) Doi Suthep #3 (including A. caudata and various saprobic fungi), (c) Doi Pee Pan Nam (including C. exigua , two ECM fungi and a saprobic Marasmius ). Letters in brackets denote significant differences between species for both 13 C (first letter) and 15 N (second letter), according to pairwise Mann-Whitney tests ( P < 0.01 at least); bars indicate standard deviations.

Page 12 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

a 35 a 30 a 25 b C/N 20 b b 15 a a a a a a 10

0 inerme gerardii milax sp. Lactarius Canthium S rotunda montana osenbergia Lactarius B Aphyllorchis Lactarius sp. Castanopsis autrozonarius Tomentella sp. Cortinarius sp. acuminatissima Thelephora sp. b 35 b b 30 a, b, c 25 20 c a C/N a a a 15 a 10

0 ursii ursii Ludisia discolor k k acaranga aprophyte tereum sp. caudata communis Lasianthus S Commelina M S Collybia sp. Aphyllorchis Acanthaceae

c 30 a a 25 a a 20 15 c c b C/N 10 b b 5 0 elatior sultani mpatiens milax milax sp. Aspidista I S montana exigua dyerinaus Cheirostylis trobilenthes Laccaria sp. arasmius sp. S M Cephalanthera Thelephora sp.

C/NFigure ratio 9 values of the three mycoheterotrophic orchids and other green orchids C/N ratio values of the three mycoheterotrophic orchids and other green orchids C/N ratio values of green plants, (white bars), mycoheterotrophic plants (grey bars) and fungi (black bars) from three sites: Doi Suthep #2 ( (a) , including A. montana and various ectomycorrhizal (ECM) fungi), Doi Suthep #3 ( (b) , including A. caudata , various saprobic fungi ), Doi Pee Pan Nam ( (c) , including C. exigua , two ECM fungi and a saprobic Marasmius ). Letters denote significant differences between species, according to pairwise Mann-Whitney tests ( P < 0.01 at minimum); bars indicate the standard deviation.

Page 13 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

tain forests [28]. C. exigua blooms during the dry season rbcL1F (5'-ATGTCACCACAAACAGAAAC-3') and rbcL (April), whereas the two Aphyllorchis spp. bloom during 1367R (5'-CTTCCAAATTTCACAAGCAGCA-3'); and (iii) the rainy season (July to August). All roots were harvested trnS-G using a primer on trnS (5'-GCCGCTTTAGTCCACT- from large populations at the beginning of their flowering CAGC-3') and the other on trnG (5'-GAACGAAT- period in 2006 and 2007, with the authorisation of the CACACTTTTACCAC-3'). These loci were also amplified in National Council for Research of Thailand. Samples were other Neottieae, such as Cephalanthera exigua , C. damaso- collected from 10 different sampling sites, separated by nium Druce, C. longifolia (L.) Fritsch, C. rubra (L.) Rich., 500 m to 1000 km in diverse parts of Thailand (North- Epipactis helleborine (L.) Crantz, E. muelleri Godfery, E. West, Central and South-East) with different forest types fageticola (C.E.Hermos.) Devillers-Tersch. & Devillers, E. (evergreen, pine-oak or dry dipterocarpacean forest) – see fibri Scappat. & Robatsch, E. palustris Crantz, E. flava Sei- details and site names in Table 1. denf., E. microphylla Sieber. ex Nyman, Neottia ovata Bluff & Fingerh., N. nidus-avis (L.) Rich., Limodorum abortivum Sampling for molecular analysis (L.) Sw., and Thaia saprophytica Seidenf. (Additional file We harvested three to six independent root fragments (> 1). Tropidia curculigoides Lindl. was sequenced as an out- 3 cm in length) using a protocol that allows plant survival group. Sequencing and sequence editing was performed (careful approach to plant roots by digging from one side as in Roy et al. [8] and corrected sequences (or consensus and, after sampling, refilling of the hole with the same soil sequences for similar clones) were deposited in GenBank without direct rhizome disturbance [8]). We discarded [75]. roots specialised in starch accumulation (often occurring in Neottieae [31]) and roots showing infections or symp- Fungal identification and phylogenetic analyses toms of decay. Within 2 h after harvesting, the remaining In order to identify fungi, a BLAST search for similar fun- roots were carefully washed with water to eliminate soil gal sequences was conducted [76] using GenBank [75]. particles, surface-sterilised using a solution of sodium Two phylogenetic analyses were conducted, in order to (i) hypochloride (2% v/v) and Tween 80 (5% w/v) for 10 s, study the phylogenetic position of Aphyllorchis spp. using and rinsed three times in sterile distilled water. Roots were a concatenation of ITS, rbcL and trnS-G sequences, and (ii) then enveloped in paper and stored in silica gel. Next, 1 to refine the phylogenetic positions of the many Russu- mm-long sections were sampled every centimetre on the laceae found in this study using ITS sequences (alignment roots, and their colonisation was checked under the and analysis were not possible for Thelephoraceae, microscope using the neighbouring root section (3 to 15 because of too much variation in their ITS). Sequences of colonised samples were recovered per plant). To identify Neottieae and Russulaceae available in GenBank were directly the fungi forming pelotons (intracellular hyphal downloaded and aligned together with ours using Clus- coils produced by orchid mycorrhizal fungi), pelotons talW [77], and then corrected by eye. Considering the high were isolated under a microscope according to Rasmussen number of species of Russulaceae in GenBank, we used [31] on A. montana individuals AMD6.1 and AMD7.1 only species recorded from Thailand and species recov- from Doi Suthep #2 (Table 1 and Additional file 2). For ered when using BLAST for our sequences; the Russu- 10 root sections per individual, 12 pelotons were recov- laceae tree was not rooted. For Neottieae, Tropidia ered and pooled per section (2 × 10 = 20 peloton pools in polystachya, Nervilia shinensis and Vanilla planifolia were all). chosen as outgroups. The phylogeny was computed by maximum likelihood with PhyML v2.4.4 [78]. For this Molecular investigations analysis, a general time-reversible (GTR) model of DNA DNA extraction and PCR amplification of fungal ITS of substitution was used [79,80], involving unequal base fre- ribosomal DNA were performed as in Selosse et al. [5] on quencies and six types of substitution. This model of DNA root fragments and peloton pools. Whenever PCR failed, substitution was chosen using a series of hierarchical like- we tentatively amplified (i) the large mitochondrial ribos- lihood-ratio tests in Modeltest 3.7 [81]. Base frequencies omal subunit gene (LrDNA) as in Roy et al. [8], and (ii) were estimated, and 10,000 bootstrap replicates were per- the 5' part of the 28S rDNA, using the primers Lr0r and Lr5 formed. Phylogenetic trees were visualised using Figtree as in Roy et al. [8]. Some PCR products with multiple 1.1.2 [82]. bands were cloned as in Roy et al. [8], and at least five clones per individual were recovered. Before sequencing, Isotopic sampling and analysis RFLP, using EcoR I+ Sac I and Hind III, as in Selosse et al. [5] Sampling for isotopic studies was conducted at three dif- was investigated to avoid repetitive sequencing of the ferent sites (Doi Suthep #2, Doi Suthep #3 and Doi Pee same ITS. To investigate the phylogenetic position of the Pan Nam; Table 1). At each site, we harvested n = 5 sam- investigated orchid species, we amplified (conditions in ples for aerial parts of MH orchids, leaves of four Selosse et al. [5]) and sequenced (i) the plant ITS, using autotrophic species, and fruitbodies of up to six basidio- the plant-specific primer ITS1P; (ii) rbcL using primer mycetes species fruiting at sampling time (prioritising

Page 14 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

ECM species). All leaves were collected in close vicinity, at Authors' contributions the same apparent light level and the same distance from This article is part of MR's PhD thesis. M-AS and SV the ground (less than 0.5 m) as orchids. When available, designed the research; SW and MR performed the sam- other terrestrial green orchids were collected ( Ludisia dis- pling; MR and AS performed the molecular research; MR, color at Doi Suthep #3 and Cheirostylis montana at Doi Pee M-AS and FR analysed the data; MR and M-AS wrote the Pan Nam). Samples were dried at 65°C for 72 h and han- paper. All authors have read and approved the final man- dled as in Tedersoo et al. [21] to measure total N, C/N uscript. ratio and abundances of 13 C and 15 N. Isotope abundances are expressed in 13 C and 15 N values in parts per thou- Additional material sand relative to international standards V-PDB and atmos- 13 15 pheric N 2: C or N = (R sample /R standard - 1) × 1000 Additional file 1 13 12 15 (‰), where R is the molar ratio, that is, C/ C or N/ Table S1 14 . Origin and accession numbers of orchids collected for the Neot- N. The standard deviation of the replicated standard tieae phylogeny. Bold numbers indicate the sequences obtained in this 13 samples ( n = 13) was 0.031‰ for C and 0.237‰ for study, the others were retrieved from GenBank. 15 N. Total N, C/N ratio, 13 C and 15 N values were com- Click here for file pared independently between species at each site by pair- [http://www.biomedcentral.com/content/supplementary/1741- wise Mann-Whitney tests using Minitab™. Thus, groups of 7007-7-51-S1.xls] species were delimited for each variable and the Kruskal- Additional file 2 Wallis test was performed, using these groups as a factor Table S2 . Fungi identified from , and in order to study the validity of these groups more pre- A. montana A. caudata C. exigua mycorrhizae. Putative species were delineated on a 97% internal tran- cisely. scribed spacer similarity threshold; whenever several species belong to the same taxon, their name includes a number to distinguish them, e.g. Rus- Fungal community analysis sulaceae sp. #3. Putative ecologies: ECM, ectomycorrhizal fungus; E: To infer species from ITS sequences, we applied a thresh- endophyte ( sensu Julou et al. [41]); R: rhizoctonia; S: saprobe. ITS: old of 97.0% sequence identity over the whole ITS region; Internal transcribed spacer; 28S: large ribosomal DNA subunit (28S); LrDNA: mitochondrial ribosomal DNA. Small letters refer to the DNA although there is no universally applicable threshold [83], source used for identification, and upper numbers to the number of iden- this is in agreement with our previous studies [8,21]. tical sequences retrieved from the same individual: d, direct PCR from Sequences were aligned using Bioedit and a similarity root DNA; c, cloned PCR product from root DNA, r, inferred from RFLP matrix was calculated. The frequency (p i) of each putative profile; p, PCR from peloton DNA. ECM species among individuals and within populations Click here for file was calculated to establish a Shannon diversity index and [http://www.biomedcentral.com/content/supplementary/1741- 7007-7-51-S2.xls] a Simpson diversity index. Indices were compared between individuals by the pairwise Mann-Whitney test. To account for our variable sampling effort among orchid species, rarefaction curves were simulated 5,000 times Acknowledgements using analytic rarefaction 1.3 [84] on two datasets: one The authors warmly thank C Debain, M Hossaert, M-P Dubois, A Meekiij- pooling all populations for each species, and the other jaroenroj and P Feldmann for help in launching this research programme, as separating each population and calculating a mean value well as C Shtultz and three anonymous referees for helpful comments on for each species. For a more qualitative analysis, fungal this article and David Marsh for editing the article. We thank the Queen communities at the individual level were compared Sirikit Botanical Garden and particularly Dr W Nanakorn for supporting within and between species by building similarity this project, K Srimuang, P Panyachan and P-H Fabre for their help during matrixes with Primer 5.2.9 [85] using the Bray-Curtis sim- sampling. Molecular data used in this work were produced through molec- ilarity index. Two matrixes were computed by grouping ular genetic analysis technical faciliti es of the 'Service des Marqueurs Géné- tiques en Ecologie' at the 'Centre d'Ecologie Fonctionnelle et Evolutive' and fungal species into families (because no species or of the IFR119 'Montpellier Environnement Biodiversité'. M-AS is funded by sequence was common between orchid populations or the Centre National de la Recherche Scientifique and the Société Française species, see below). DCA was performed with these d'Orchidophilie. matrixes, using population, forest type, geographical origin and species as factors. References 1. Leake JR: Myco-heterotroph/epiparasitic plant interactions with ectomycorrhizal and arbuscular mycorrhizal fungi. Curr Abbreviations Opin Plant Biol 2004, 7: 422-428. AC: Aphyllorchis caudata; AM: Aphyllorchis montana; CE: 2. Selosse MA, Roy M: Green plants that feed on fungi: facts and Cephalanthera exigua; ECM: ectomycorrhiza or ectomycor- questions about mixotrophy. Trends Plant Sci 2009, 14(2): 64-70. 3. Smith SE, Read DJ: Mycorrhizal symbiosis 3rd edition. London, UK: Aca- rhizal; DCA: detrended component analysis; GTR: general demic Press; 2008. time reversible (model); ITS: internal transcribed spacer of 4. Leake JR: The biology of myco-heterotrophic (Saprophytic) the ribosomal DNA; MH: mycoheterotroph. Plants. New Phytol 1994, 127: 171-216.

Page 15 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

5. Selosse MA, Weiss M, Jany JL, Tillier A: Communities and popula- 28. Seidenfaden G: Orchid genera in Thailand VI: Neottioideae Lindl Dansk tions of sebacinoid basidiomycetes associated with the Botanisk Forening; 1978. achlorophyllous orchid Neottia nidus-avis (L.) LCM Rich. and 29. Suarez JP, Weiss M, Abele A, Garnica S, Oberwinkler F, Kottke I: neighbouring tree ectomycorrhizae. Mol Ecol 2002, Diverse tulasnelloid fungi form mycorrhizas with epiphytic 11: 1831-1844. orchids in an Andean cloud forest. Mycol Res 2006, 6. Taylor DL, Bruns TD, Leake JR, Read DJ: Mycorrhizal specificity 110: 1257-1270. and function in myco-heterotrophic plants. In Mycorrhizal Ecol- 30. Otero JT, Flanagan NS, Herre EA, Ackerman JD, Bayman P: Wide- ogy Volume 157 . Edited by: Sanders I, van der Hijden M. Berlin, Ger- spread mycorrhizal specificity correlates to mycorrhizal many: Springer; 2002:375-413. function in the neotropical, epiphytic orchid Ionopsis utricu- 7. Taylor DL, Bruns TD, Szaro TM, Hodges SA: Divergence in myc- larioides (Orchidaceae). Am J Bot 2007, 94: 1944-1950. orrhizal specialization within Hexalectris spicata 31. Rasmussen HN: Terrestrial orchids – from seed to mycotrophic plant (Orchidaceae), a nonphotosynthetic desert orchid. Am J Bot Cambridge, UK: Cambridge University Press; 1995. 2003, 90: 1168-1179. 32. Alexander IJ, Lee SS: Mycorrhizas and ecosystem processes in 8. Roy M, Yagame T, Yamato M, Iwase K, Heinz C, Faccio A, Bonfante tropical rain forest: implications for diversity. In Biotic Interac- P, Selosse MA: Ectomycorrhizal Inocybe species associate with tions in the Tropics: Their Role in the Maintenance of Species Diversity the mycoheterotrophic orchid Epipogium aphyllum but not Cambridge, UK: Cambridge University Press; 2005:165-203. with its asexual propagules. Ann Bot 2009. published online. 33. Weiss M, Selosse MA, Rexer KH, Urban A, Oberwinkler F: Sebaci- 9. Dearnaley JDW: Further advances in orchid mycorrhizal nales: a hitherto overlooked cosm of heterobasidiomycetes research. Mycorrhiza 2007, 17: 475-486. with a broad mycorrhizal potential. Mycol Res 2004, 10. Bidartondo MI: The evolutionary ecology of myco-heterotro- 108: 1003-1010. phy. New Phytol 2005, 167: 335-352. 34. Athipunyakom P: Mycorrhizal fungi of terrestrial orchids: isola- 11. Bidartondo MI, Redecker D, Hijri I, Wiemken A, Bruns TD, tion, identification and symbiotic germination. In PhD thesis Dominguez L, Sersic A, Leake JR, Read DJ: Epiparasitic plants spe- Thailand: Kasetsart University; 2004. cialized on arbuscular mycorrhizal fungi. Nature 2002, 35. Pridgeon A, Cribb PJ, Chase MM: Genera orchidacearum: Epidendroidae 419: 389-392. Volume 4 . Oxford, UK: Oxford University Press; 2008. 12. Merckx V, Chatrou LW, Lemaire B, Sainge MN, Huysmans S, Smets 36. Abadie JC, Puttsepp U, Gebauer G, Faccio A, Bonfante P, Selosse MA: EF: Diversification of myco-heterotrophic angiosperms: evi- Cephalanthera longifolia (Neottieae, Orchidaceae) is mix- dence from Burmanniaceae. BMC Evol Biol 2008, 8: 1-16. otrophic: a comparative study between green and nonphoto- 13. Sainge MN, Franke T: A new species of Afrothismia (Burman- synthetic individuals. Can J Bot 2006, 84: 1462-1477. niaceae) from Cameroon. Nord J Bot 2004, 23: 299-303. 37. Taylor DL, Bruns TD: Independent, specialized invasions of 14. Yamato M, Yagame T, Suzuki A, Iwase K: Isolation and identifica- ectomycorrhizal mutualism by two nonphotosynthetic tion of mycorrhizal fungi associating with an achlorophyllous orchids. Proc Natl Acad Sci USA 1997, 94: 4510-4515. plant, Epipogium roseum (Orchidaceae). Mycoscience 2005, 38. Girlanda M, Selosse MA, Cafasso D, Brilli F, Delfine S, Fabbian R, Ghi- 46: 73-77. gnone S, Pinelli P, Segreto R, Loreto F, Cozzolino S, Perotto S: Inef- 15. Ogura-Tsujita Y, Yukawa T: High mycorrhizal specificity in a ficient photosynthesis in the Mediterranean orchid widespread mycoheterotrophic plant, Eulophia zollingeri Limodorum abortivum is mirrored by specific association to (Orchidaceae). Am J Bot 2008, 95: 93-97. ectomycorrhizal Russulaceae. Mol Ecol 2006, 15: 491-504. 16. Ogura-Tsujita Y, Gebauer G, Hashimoto T, Umata H, Yukawa T: Evi- 39. Selosse MA, Bauer R, Moyersoen B: Basal hymenomycetes dence for novel and specialized mycorrhizal parasitism: the belonging to the Sebacinaceae are ectomycorrhizal on tem- orchid Gastrodia confusa gains carbon from saprotrophic perate deciduous trees. New Phytol 2002, 155: 183-195. Mycena . Proc R Soc Lond B Biol Sci 2009, 276: 761-767. 40. Dressler RL: The Neottieae in orchid classification. Lindleyana 17. Martos F, Dulormne M, Pailler T, Bonfante P, Faccio A, Fournel J, 1990, 5: 102-109. Dubois MP, Selosse MA: Challenging ecological rules drawn 41. Julou T, Burghardt B, Gebauer G, Berveiller D, Damesin C, Selosse from temperate regions: some tropical achlorophyllous MA: Mixotrophy in orchids: insights from a comparative orchids receive carbon from saprobic fungi. New Phytol 2009 in study of green individuals and nonphotosynthetic individuals press. of Cephalanthera damasonium . New Phytol 2005, 166: 639-653. 18. Dawson TE, Mambelli S, Plamboeck AH, Templer PH, Tu KP: Stable 42. Bidartondo MI, Burghardt B, Gebauer G, Bruns TD, Read DJ: Chang- isotopes in plant ecology. Ann Rev Ecol Syst 2002, 33: 507-559. ing partners in the dark: isotopic and molecular evidence of 19. Post DM: Using stable isotopes to estimate trophic position: ectomycorrhizal liaisons between forest orchids and trees. models, methods, and assumptions. Ecology 2002, 83: 703-718. Proc R S Lond B Biol Sci 2004, 271: 1799-1806. 20. Trudell SA, Rygiewicz PT, Edmonds RL: Nitrogen and carbon sta- 43. Selosse MA, Faccio A, Scappaticci G, Bonfante P: Chlorophyllous ble isotope abundances support the myco-heterotrophic and achlorophyllous specimens of Epipactis microphylla nature and host-specificity of certain achlorophyllous plants. (Neottieae, Orchidaceae) are associated with ectomycor- New Phytol 2003, 160: 391-401. rhizal septomycetes, including truffles. Microb Ecol 2004, 21. Tedersoo L, Pellet P, Koljalg U, Selosse MA: Parallel evolutionary 47: 416-426. paths to mycoheterotrophy in understorey Ericaceae and 44. Bidartondo MI, Read DJ: Fungal specificity bottlenecks during Orchidaceae: ecological evidence for mixotrophy in orchid germination and development. Mol Ecol 2008, Pyroleae. Oecologia 2007, 151: 206-217. 17: 3707-3716. 22. Mayor JR, A SE, Henkel TW: Elucidating the nutritional dynam- 45. Lee SL, Alexander IJ, Watling R: Ectomycorrhizas and putative ics of fungi using stable isotopes. Ecol Lett 2009, 12: 171-183. ectomycorrhizal fungi of Shorea leprosula Miq. (Dipterocar- 23. Boström B, Comstedt D, Ekblad A: Isotope fractionation and 13- paceae). Mycorrhiza 1997, 7: 63-81. C enrichment in soil profiles during the decomposition of soil 46. Riviere T, Diedhiou AG, Diabate M, Senthilarasu G, Natarajan K, Ver- organic matter. Oecologia 2007, 153: 89-98. beken A, Buyck B, Dreyfus B, Bena G, Ba AM: Genetic diversity of 24. Zimmer K, Meyer C, Gebauer G: The ectomycorrhizal specialist ectomycorrhizal Basidiomycetes from African and Indian orchid Corallorhiza trifida is a partial myco-heterotroph. New tropical rain forests. Mycorrhiza 2007, 17: 415-428. Phytol 2008, 178: 395-400. 47. Yuwa-Amornpitak T, Vichitsoothonkul T, Tanticharoen M, Cheevad- 25. McKendrick SL, Leake JR, Taylor DL, Read DJ: Symbiotic germina- hanarak S, Ratchadawong S: Diversity of Ectomycorrhizal fungi tion and development of the myco-heterotrophic orchid on Dipterocarpaceae in Thailand. J Biol Sci 2006, 6: 1059-1064. Neottia nidus-avis in nature and its requirement for locally 48. Moyersoen B: Pakaraimaea dipterocarpacea is ectomycor- distributed Sebacina spp. New Phytol 2002, 154: 233-247. rhizal, indicating an ancient Gondwanaland origin for the 26. Selosse MA, Richard F, He XH, Simard SW: Mycorrhizal networks: ectomycorrhizal habit in Dipterocarpaceae. New Phytol 2006, des liaisons dangereuses? Trends Ecol Evol 2006, 21: 621-628. 172: 753-762. 27. Roberts DL, Solow AR: The effect of the convention on inter- 49. Ouanphanivanh N, Merenyi Z, Orczan AK, Bratek Z, Szigeti Z, Illyes national trade in endangered species on scientific collec- Z: Could orchids indicate truffle habitats? Mycorrhizal asso- tions. Proc R Soc Lond B Biol Sci 2008, 275: 987-989. ciation between orchids and truffles. Acta Biol Szeged 2008, 52: 229-232.

Page 16 of 17 (page number not for citation purposes) BMC Biology 2009, 7:51 http://www.biomedcentral.com/1741-7007/7/51

50. Yamato M, Iwase K: Introduction of asymbiotically propagated 72. Bronstein JL: The exploitation of mutualisms. Ecol Lett 2001, seedlings of Cephalanthera falcata (Orchidaceae) into natu- 4: 277-287. ral habitat and investigation of colonized mycorrhizal fungi. 73. Ronce O, Kirkpatrick M: When sources become sinks: migra- Ecol Res 2008, 23: 329-337. tional meltdown in heterogeneous habitats. Evolution 2001, 51. Umata H: In vitro germination of Erythrorchis ochobiensis 55: 1520-1531. (Orchidaceae) in the presence of Lyophyllum shimeji , an ecto- 74. Lanchier N, Neuhauser C: A spatially explicit model for compe- mycorrhizal fungus. Mycoscience 1997, 38: 355-357. tition among specialists and generalists in a heterogeneous 52. Yagame T, Yamato M, Suzuki A, Iwase K: Ceratobasidiaceae myc- environment. Ann Appl Probab 2006, 16: 1385-1410. orrhizal fungi isolated from nonphotosynthetic orchid 75. National Center for Biotechnology Information [http:// Chamaegastrodia sikokiana . Mycorrhiza 2008, 18: 97-101. www.ncbi.nlm.nih.gov ] 53. Terashita T, Chuman S: Fungi inhabiting wild orchids in Japan 76. Altschul SF, Madden TL, Schaffer AA, Zhang JH, Zhang Z, Miller W, (IV). Armillaria tabescens , a new symbiont of Galeola septen- Lipman DJ: Gapped BLAST and PSI-BLAST: a new generation trionalis . Trans Mycol Soc Jpn 1987, 28: 145-154. of protein database search programs. Nucleic Acids Res 1997, 54. Umata H: Seed germination of Galeola altissima , an achloro- 25: 3389-3402. phyllous orchid, with aphyllophorales fungi. Mycoscience 1995, 77. Thompson JD, Higgins DG, Gibson TJ: CLUSTAL W: improving 36: 369-372. the sensitivity of progressive multiple sequence alignment 55. Umata H: A new biological function of Shiitake mushroom, through sequence weighting, position-specific gap penalties Lentinula elodes , in a mycoheterotrophic orchid , Erythrorchis and weight matrix choice. Nucleic Acids Res 1994, 22: 4673-4680. ochobiensis . Mycoscience 1998, 39: 85-88. 78. Guindon S, Gascuel O: A simple, fast, and accurate algorithm 56. McKendrick SL, Leake JR, Read DJ: Symbiotic germination and to estimate large phylogenies by maximum likelihood. Syst development of myco-heterotrophic plants in nature: trans- Biol 2003, 52: 696-704. fer of carbon from ectomycorrhizal Salix repens and Betula 79. Lanave C, Preparata G, Saccone C, Serio G: A new method for cal- pendula to the orchid Corallorhiza trifida through shared culating evolutionary substitution rates. J Mol Evol 1984, hyphal connections. New Phytol 2000, 145: 539-548. 20: 86-93. 57. Zeller B, Brechet C, Maurice JP, Le Tacon F: C-13 and N-15 iso- 80. Rodriguez FJ, Olivier JL, Marin A, Medina JR: The general stochas- topic fractionation in trees, soils and fungi in a natural forest tic model of DNA substitution. J Theor Biol 1990, 142: 485-501. stand and a Norway spruce plantation. Ann Forest Sci 2007, 81. Posada D, Crandall KA: MODELTEST: testing the model of 64: 419-429. DNA substitution. Bioinformatics 1998, 14: 817-818. 58. McGuire KL: Common ectomycorrhizal networks may main- 82. Figtree 1.1.2 [http://tree.bio.ed.ac.uk/ ] tain monodominance in a tropical rain forest. Ecology 2007, 83. Nillsson RH, Kristiansson E, Ryberg M, Hallenberg N, Larsson kH: 88: 567-574. ITS variability in the Kingdom Fungi as expressed in the 59. Gebauer G, Meyer M: N-15 and C-13 natural abundance of international sequence databases and its implications for autotrophic and mycoheterotrophic orchids provides insight molecular species identification. Evol Bioinform Online 2008, into nitrogen and carbon gain from fungal association. New 4: 193-201. Phytol 2003, 160: 209-223. 84. Holland SM: Analytic Rarefaction Version 1.3. 2008 [http:// 60. Phillips DL, Gregg JW: Uncertainty in source partitioning using www.uga.edu/strata/software/ ]. stable isotopes. Oecologia 2001, 128: 304. 85. Clarke KR, Gorley RN: Primer v5.2.9: User Manual: Tutorial, Primer-E. 61. Bougoure JJ, Bougoure DS, Cairney JWG, Dearnaley JDW: ITS- Plymouth UK 2001. RFLP and sequence analysis of endophytes from Acianthus, Caladenia and Pterostylis (Orchidaceae) in southeastern Queensland. Mycol Res 2005, 109: 452-460. 62. Tedersoo L, Suvi T, Beaver K, Koljalg U: Ectomycorrhizal fungi of the Seychelles: diversity patterns and host shifts from the native Vateriopsis seychellarum (Dipterocarpaceae) and Intsia bijuga (Caesalpiniaceae) to the introduced Eucalyptus robusta (Myrtaceae), but not Pinus caribea (Pinaceae). New Phytol 2007, 175: 321-333. 63. Taylor DL, Bruns TD, Hodges SA: Evidence for mycorrhizal races in a cheating orchid. Proc R Soc Lond B Biol Sci 2004, 271: 35-43. 64. Franke T, Beenken L, Doring M, Kocyan A, Agerer R: Arbuscular mycorrhizal fungi of the Glomus-group A lineage (Glomer- ales; Glomeromycota) detected in myco-heterotrophic plants from tropical Africa. Mycol Prog 2006, 5: 24-31. 65. Dearnaley JDW, Le Brocque AF: Endophytic fungi associated with Australian orchids. Australas Plant Conserv 2006, 15: 7-9. 66. Otero JT, Ackerman JD, Bayman P: Diversity and host specificity of endophytic Rhizoctonia -like fungi from tropical orchids. Am J Bot 2002, 89: 1852-1858. 67. Suarez JP, Weiss M, Abele A, Oberwinkler F, Kottke I: Members of Sebacinales subgroup B form mycorrhizae with epiphytic Publish with Bio Med Central and every orchids in a neotropical mountain rain forest. Mycol Prog 2008, 7: 75-85. scientist can read your work free of charge 68. Arnold AE, Henk DA, Eells RL, Lutzoni F, Vilgalys R: Diversity and "BioMed Central will be the most significant development for phylogenetic affinities of foliar fungal endophytes in loblolly disseminating the results of biomedical research in our lifetime." pine inferred by culturing and environmental PCR. Mycologia 2007, 99: 185-206. Sir Paul Nurse, Cancer Research UK 69. Novotny V, Basset Y, Miller SE, Weiblen GD, Bremer B, Cizek L, Your research papers will be: Drozd P: Low host specificity of herbivorous insects in a tropical forest. Nature 2002, 416: 841-844. available free of charge to the entire biomedical community 70. Bruns TD, Bidartondo MI, Taylor DL: Host specificity in ectomy- peer reviewed and published immediately upon acceptance corrhizal communities: What do the exceptions tell us? Integr Comp Biol 2002, 42: 352-359. cited in PubMed and archived on PubMed Central 71. Cameron DD, Johnson I, Read DJ, Leake JR: Giving and receiving: yours — you keep the copyright measuring the carbon cost of mycorrhizas in the green orchid, Goodyera repens . New Phytol 2008, 180: 176-184. Submit your manuscript here: Bio Med central http://www.biomedcentral.com/info/publishing_adv.asp

Page 17 of 17 (page number not for citation purposes) 79.(1*= 79.(1*= = = _8_= = ]= _= = ,**3= 7**3= 51&398= 9-&9= +**)= 43= +:3,.a= +&(98= &3)= 6:*89.438= &'4:9=2.=49745->_ =7*3)8=.3= 1&39= (.*3(*8 =+.,a=/380-_= = &=2.=49745-.*=(-*?=1*8=47(-.)S*8=&=S9S=)S(4:;*79*=7S(*22*39`=349&22*39=,7B(*=A=1&= 57S8*3(*= ).3).;.):8= )S54:7;:8= )*= (-14745->11*= *9= 8:7;.;&39= A= 1B,*= &):19*= &1'.348_= 0*5*3)&39`=)*8=*85T(*8=8&38=&1'.348=8*=8439=&:88.=7S;S1S*8=2.=49745-*8`=(422*= -*.7489>1.8= 2439&3&= (+_ &79.(1*= = 6:.= 3*89= 5&8= 343= 51:8= 574(-*= 5->14,S3S9.6:*2*39= ):3*= *85T(*= 2>(4-S9S749745-*_== 0*9= &79.(1*= &= 54:7= ':9= )*= +&.7*= 1*= 54.39= 8:7= 348= (433&.88&3(*8= &:= 8:/*9= )*= 1&= 2.=49745-.*= *9= 548*= 1*8= 6:*89.438= &:=6:*11*8= /&.= *88&>S= )*= 7S543)7*= )&38= 2&= 9-T8*= 349&22*39= *3= (-*7(-&39= )*8= 2.=49745-*8$= .384:5O433S*8= (+_ &79.(1*= `= *3= (-*7(-&39= A= (4257*3)7*= 1S;41:9.43= )*= (*= 7S,.2*= &1.2*39&.7*= (+_ &79.(1*8= = *9= = 4:= *3= (-*7(-&39= A= S9:).*7=843=+43(9.433*2*39=*9=8*8=;&7.&9.438=85&9.489*2547*11*8= (+_ &79.(1*= _= 71=8&,.88&.9=)*=+&.7*=:3=S9&9=)*8=1.*:==*9=)*=).++:8*7=51:8=1&7,*2*39=(*8=7S8:19&98`=*9=(*= +29=:3=8:((T8=5:.86:*=(*9=&79.(1*=+.9=1&=57*2.T7*=5&,*=)*= 7*3)8=.3= 1&39= (.*3(*8 _= = = = This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy

Opinion

Green plants that feed on fungi: facts and questions about mixotrophy

Marc-Andre ´ Selosse and Me ´ lanie Roy

Centre d’Ecologie Fonctionnelle et Evolutive (CNRS, UMR 5175), Equipe Interactions Biotiques, 1919 Route de Mende, 34293 Montpellier ce´dex 5, France

Several green, photosynthetic plants in orchids and Furthermore, we suggest that these tools now allow the Ericaceae were recently found to recover carbon from question of the general relevance and ecological import- the mycorrhizal fungi associated with their roots, a dual ance of mixotrophy to be addressed. We propose some nutritional capability called mixotrophy. The physiologi- research hypotheses on its ecological meaning and evol- cal and cellular processes allowing carbon gain from the ution. fungus are not well understood. We believe that this phenomenon is overlooked and propose several land The diversity of mixotrophic eukaryotes plant families and ecosystems that should be investi- C availability is a limiting factor for many organisms. In gated for possible mixotrophy. We speculate that mix- autotrophs, light availability might also limit C nutrition, otrophy allowed, in some lineages, the evolution of and many photosynthetic lineages obtain additional C in a heterotrophic plants, that is, non-photosynthetic plants heterotrophic manner. The resulting mixotrophy was long that obtain their carbon from organic compounds. More- ago reported and studied for planktonic algae that show a over, the amount of carbon gained from the fungus wide range of variation, from autotrophs with facultative varies from one site to another in mixotrophs. Drawing heterotrophy to heterotrophs that are facultatively auto- a parallel with mixotrophy in planktonic algae, we pro- trophic [8] . In some algae, C is obtained from dissolved pose some hypotheses that could account for this. organic matter: Chlamydomonas reinhardtii grown on acetate derives 50% of its C from this source [9] . Most A newly discovered form of plant nutrition often, mixotrophy in algae is achieved by phagocytosis of A recent breakthrough in the understanding of plant nutri- small planktonic prey, a strategy found in various, phylo- tion is the discovery that some green plants from temper- genetically unrelated eukaryotic taxa, such as Chlorarach- ate forests not only perform photosynthesis but also obtain niophyta, Dinophyta, Ciliates, Haptophyta and additional carbon (C) from their symbiotic fungi. This Cryptophyta [10] . Planktonic algae achieve up to 95% of second C source results from the exploitation of existing the bacterivory in the superficial ocean layer [11] . In these mycorrhizal symbioses that link soil fungi with the roots of algal lineages, phagocytosis is a conserved ancestral trait circa 90% of land plants [1] . In the typical mycorrhiza, the that previously allowed nutrition of heterotrophic ances- fungus provides mineral resources, which are collected by tors and, at a certain time, allowed the engulfment of free- the fungal soil mycelium, and receives plant photo- living autotrophic cells that evolved into plastids [10] . synthates as a reward. However, some green orchids (in Epipactis , Cephalanthera , Plantanthera and other genera Glossary [2 –5] ) and, more recently, some small green perennial shrubs from the Ericaceae family (in the genera Pyrola , Autotroph : an organism that is able to use atmospheric CO 2 as its sole carbon source, for example by way of photosynthesis. Orthilia and Chimaphila , collectively referred to as pyr- Hemiparasite : a plant, such as mistletoe, that although it is capable of oloids [5,6] ) have been shown to receive variable amounts performing photosynthesis, lives parasitically on other plants, from which it of C from their mycorrhizal fungi. obtains mineral nutrients and water. In some cases, mixotrophy occurs: the hemiparasites obtain carbon compounds from the sap of the host. Plants using fungal C are not new to science: several Mixotroph : an autotrophic organism that combines its photosynthesis and a non-green plants are already known to rely solely on C partial heterotrophy as carbon sources (synonyms: hemi-autotroph or partial from their mycorrhizal fungi ( Box 1 ) [7] . These plants are mycoheterotroph). Indeed, a continuum from autotrophic to fully heterotrophic organisms exists in nature. called mycoheterotrophic because mycorrhizal fungi sup- Mycoheterotroph : a non-photosynthetic, non-chlorophyllous plant that ob- port their heterotrophy. Indeed, mixotrophic orchids and tains not only minerals but also carbon from its mycorrhizal fungus. Mycoheterotrophs were previously called ‘saprophytic plants’ or ‘saprobic pyroloids are phylogenetically related to some mycoheter- plants’ (see Box 1). otrophic species, as discussed below. The true break- Mycorrhiza : a symbiotic association between a soil fungus and a plant root. It through in plant nutrition reported here is the finding of is usually a mutualism in which plant photosynthates are exchanged for mineral resources acquired by the fungus from the soil, but some plants, such a dual nutrition in plants, partly auto- and partly hetero- as mixo- and mycoheterotrophs, can reverse the carbon flow. trophic. Here, we summarize how two investigation tools Phagocytosis : a cellular process by which structures are engulfed and digested (stable isotopes and molecular identification of the fungi) in a eukaryotic cell, for example hyphae penetrating host cells in some orchid mycorrhizae, or prey acquired by unicellular predators. have provided new insights into this phenomenon. Pyroloids : members of a tribe ( Pyroleae ) in the family Ericaceae, species of which have been recently shown to be mixotrophic. A sister clade to the two mycoheterotrophic Ericaceae tribes Monotropeae and Pterosporeae . Corresponding author: Selosse, M.-A. ( [email protected] ).

Opinion Trends in Plant Science Vol.14 No.2

Box 1. Mycoheterotrophy, a heterotrophic lifestyle that has frequently arisen in plant evolution

More than 400 heterotrophic plants receive all their nutrients, including carbon (C), from their mycorrhizal fungi [7] . These plants were once considered as saprobic, but due to the fungal origin of their nutrients, they are now called mycoheterotrophs [7,46] . Much recent work has focused on the identification of their mycorrhizal fungi. In more than 30 studied species, two common features were observed: each mycoheterotrophic plant is associated specifically with a fungal clade (a genus or a sub-group within a genus) and the fungi involved are also mycorrhizal on surrounding autotrophic plants (a few tropical orchids might, however, associate with soil saprobic fungi [47] ). This high specificity between symbionts contrasts with the low specificity in mycorrhizal associations of non-mycoheterotrophic plants, where a fungal species is associated with several plant species and vice versa [1] . Mixotrophic plants show either specific [21,22] or non-specific [3 – 6,25,29] mycorrhizal associations. Reasons for mycoheterotrophic specificity remain debated, but it is often considered that an evolutionary arms race in a parasitic context can lead to specialization [46] , when a mycoheterotrophic plant evolves to parasitize the fungus and the fungus reciprocally evolves to get rid of it. Unfortunately, so far we have no direct evidence that mycoheterotrophs are detrimental to their mycorrhizal fungi, and parasitism is thus not demonstrated. For most mycoheterotrophic plants, the fungus providing C is also mycorrhizal on nearby autotrophic plants; indeed, labelling experiments demonstrated that these plants are the ultimate C source [48] , by way of hyphal links. Thus, a mycoheterotrophic plant is equivalent to a mixotrophic plant that would not rely on photosynthesis for C supply, but only on fungal C ( Figure 1 b in main text). As a result, 13 C abundance in the mycoheterotrophic plant and the associated mycorrhizal fungi is the same (see Monotropa hypopitys in Figure 1 a in main text). Although absolute values of 13 C abundance vary from one site to another, the observed difference between mycoheterotrophic and autotrophic plants from the same site is relatively stable ( 7% in d13 C [21] ). All available data indicate that shifts from mixotrophy to mycoheterotrophy are likely to have occurred in evolution. With increasing knowledge of the phylogenetic position of mycoheterotrophic plants, there is some evidence that they arose repeatedly from mixotrophic Figure I . A phylogeny of the orchid tribe Neottieae supports a series of states leading from autotrophy to mixotrophy and then to mycoheterotrophy (tree ancestors ( Figure I ). So far, this has been documented in orchids based on rbcL , ITS and trnS-G by maximum likelihood method; red asterisks [3,4,25] and pyroloids [5] , where mixotrophy has been demonstrated. indicate branches supported by >85% bootstrap values after 1000 replicates). In the series of states leading from autotrophy to mixotrophy and The most parsimonious ancestral states are represented; the following colour then to mycoheterotrophy, reversal from mixotrophy to full auto- code is used for names and branches: black, unknown trophic state; orange, trophy is suspected to occur ( [29] , see E. palustris in Figure I ). By mycoheterotrophic; blue, mixotrophic; green, autotrophic. contrast, reversal from mycoheterotrophy to mixotrophy has not been documented, perhaps because of the irreversible alteration of photosynthetic genes in mycoheterotrophs [49] .

By contrast, mixotrophy among land plants is poorly trapping devices) and occur only in a few families, whereas studied and seems to be a secondarily, recently derived the newly discovered mixotrophy in orchids and pyroloids feature [5] (Box 1 ). It was sometimes reported as ‘hemi- simply relies on mycorrhizae, a symbiosis already present autotrophy’ or ‘partial heterotrophy’, but the broader term in most land plants [1] . It is thus potentially a more ‘mixotrophy’ uniﬁes the diverse occurrences of organic C widespread strategy. For simpliﬁcation, this mycorrhizal acquisition by photosynthetic eukaryotes in all ecosystems. mixotrophy is called mixotrophy throughout this text. Moreover, unifying all mixotrophic models allows the for- mulation of hypotheses for land plants starting from the Plant carbon supply from two distinct sources better known algal mixotrophy. 14 C or 13 C labelling experiments have demonstrated that Mixotrophy has long been known in hemiparasites, that various photosynthetic plants can receive C from their is, plants that parasitize other plants but retain photosyn- fungal associates [1,16,17] . However, such methods only thetic abilities. Besides producing their own photo- report instantaneous transfers, which might not occur synthates, some hemiparasites gain organic C from host continuously or under all conditions [18,19] . Moreover, plants [12] . For example, mistletoes derive up to 63% of they do not account for the exact contribution of such their C from their host [13] . The possibility that some transfers to the plant’s C budget over the whole growing carnivorous plants derive C from their prey [14] , as a season. By contrast, natural abundance in 13 C in plant side-product of the absorption of organic nitrogen (N) biomass is a powerful tool for unravelling food sources in and phosphorus (P), has also been reported, but somewhat situ [14,20] ; food sources over the whole lifespan of overlooked [15] . However, parasitic and carnivorous plants the plant can be interpreted without experimental have evolved specialized structures (such as haustoria or disturbance. Compared to neighbouring autotrophic

65 Author's personal copy

Opinion Trends in Plant Science Vol.14 No.2

Whenever gas exchanges have been investigated, mixotrophic plants have demonstrated the ability to ﬁx atmospheric CO 2 through photosynthesis ( Figure 1b, 2 ) [4,5,22] . However, depending on the species, low light conditions [4] , low chlorophyll content [21] or low photosynthetic activity [22] limit the plants’ photosynthesis rate to a level equal to or lower than their respiration. These plants thus clearly need an additional C source for growth and reproduction. In some orchids, for example Corallorhiza triﬁda [21] and Limodorum abortivum [22] , strong reduction of leaf size and number also limits photosynthesis. As a result, a large range of variation in the level of heterotrophy (the plant’s dependence on the fungal C) can be expected. This was further measured by 13 C abundances in mixotrophs, whose biomass results from mixing a proportion ( p) of fungal C with (1- p) of photosynthetic C. Using the 13 C abundances in mycoheterotrophic and autotrophic plants as references, p was calculated to range from 0 to 85% in investigated mixotrophs [2,5,6] . As expected, this value varies among species as well as from one site to another for a given species [2,5,6] . A major limitation of these calculations is that they establish the contribution of fungal C to biomass (anabolism) but not to catabolism. Investigations on respiratory CO 2 will thus be necessary to build a global view of C metabolism in mixotrophs.

The fungal symbionts: carbon donors and recipients Molecular methods are now widespread in microbial ecology and are especially useful in identifying poorly cultivable mycorrhizal fungi. Mycologists have developed Figure 1 . Stable isotope abundance in mixotrophic and mycoheterotrophic plants. barcoding methods based on fungal rDNA, for which refer- Isotope abundance is expressed in d13 C and d15 N values (in parts per thousand) ence sequences exist in public databases [23] . Identiﬁ- relative to international standards: d13 C or d15 N = [(R(sample)/ R(standard) À 1] Â 1000, where R is the molar ratio (i.e. 13 C: 12 C or 15 N: 14 N) cation of the mycorrhizal fungi of mixotrophs is more (reproduced from Ref. [5] , with permission). (a) Stable isotope abundance in than a purely descriptive task because the putative ecology plants and mycorrhizal fungi from a boreal Estonian forest (reproduced from Ref. of these fungi provides clues to the ultimate C source that is [5] , with permission). Species include mixotrophic and autotrophic orchids (O), as well as pyroloids (P) and Ericaeae phylogenetically related to pyroloids (the being exploited. The mycorrhizal fungi that are associated autotrophic Arctostaphylos uva-ursi and the mycoheterotrophic Monotropa with pyroloids and orchids belong to diverse fungal taxa hypopitys ). Code for colours: brown, mycorrhizal fungi associated with (Ascomycetes and Basidiomycetes) that usually form surrounding trees and mixotrophs; orange, mycoheterotrophic plants; blue: mixotrophic plants; green: autotrophic plants. The fungus indicated by an arrow mycorrhizae on tree roots [3,5,6,22,24] . In some reports, (Tricholoma myomyces ) is mycorrhizal on M. hypopitys . (b) A summary of C they were simultaneously detected on surrounding trees of (orange) and N (blue) nutrient flow in mixotrophic plants and associated plants the study site [4,25] . Mixotrophic plants that live in forests (green tree) and fungi, with ranges of isotopic fractionation at the various plant – fungus interfaces (see Ref. [6] for a review). thus derive photosynthates from overstory trees to supple- ment their photosynthesis, by way of sharing fungi mutually ( Figure 1 b). This adds to the growing body of plants, mixotrophic orchids and pyroloids are enriched in evidence that coexisting plants share mycorrhizal fungi 13 C, although they perform the same C3 photosynthesis that mediate plant –plant interactions [17,18] , so that some (Figure 1 a) [2 –6,21] . The high amount of 13 C in mixotrophic plants can evolve adaptations to use the resulting plant – plants tends to be close to the amount of 13 C found in their plant links. mycorrhizal fungi, a situation reported for mycohetero- It is noteworthy that, so far, no published labelling trophic plants ( Box 1 ). 15 N abundances are also congruent experiment has directly supported the mixotrophic with the use of fungal resources ( Figure 1 a). This isotope scenario of Figure 1 b, although such experiments have becomes more concentrated along most trophic chains: already demonstrated C transfer from surrounding trees fungi that often exploit substrates derived from plant to mycoheterotrophic plants [18] . Moreover, mechanisms substrates (directly or after recycling) are richer in 15 N transferring fungal C to mixotrophs (and mycohetero- than autotrophic plants. In turn, plants feeding on fungi trophs) remain fully unknown. In orchid mycorrhizae, are richer in 15 N than their mycorrhizal fungi [2] . Due to fungal hyphae colonize the root cells and ﬁnally undergo their mixed nutrition, mixotrophic plants have 15 N abun- lysis [1] of unclear role: it can be interpreted as mobiliz- dances that lie above those measured in autotrophic plants ation of nutrients or as a simple recycling of old structures. and close to those measured in mycorrhizal fungi An appealing scenario considers it as a phagocytosis-like (Figure 1 a). process mobilizing fungal C. This well explains why the 13 C

66 Author's personal copy

Opinion Trends in Plant Science Vol.14 No.2

Figure 2 . CO 2 uptake in response to light levels (a) in Cephalanthera damasonium , a mixotrophic, green orchid species in which non-chlorophyllous variants exist that survive as mycoheterotrophs. (b) A typical green C. damasonium individual is pictured together with a non-chlorophyllous variant (on the right), which shows no CO 2 uptake due to the lack of any photosynthetic ability. Reproduced from Ref. [4] , with permission.

content of the receiving plant is similar to that of the Evolutionary predisposition to heterotrophy associated fungi, because hyphae are totally digested Mixotrophy has important evolutionary implications. As and absorbed [26] . Two facts contradict the digestion previously mentioned, mixotrophic orchids and pyroloids model, however: 15 N values differ from associated fungi are phylogenetically close to mycoheterotrophs that fully in mycoheterotrophs [26] and, although fungal hyphae also rely on fungal C ( Box 1 ) [25,29] . Available phylogenies (e.g. penetrate root cells of pyroloids, no lysis of fungi has been Figure I in Box 1 ) show that mixotrophy appeared first and observed in these species [5,24] . Thus, other transfer probably facilitated the emergence of mycoheterotrophy. mechanisms might act in the mobilization of fungal C. The later transition probably arose through a shift to An alternative scenario involves transfer of organic mol- absolute dependence on fungal C. The ecological niche of ecules from living hyphae to host cells. An intriguing mixotrophs might facilitate such transitions: mixotrophs possibility is that C and N transfers are linked together: often grow in young forests with loose canopy, where the C although mycorrhizal fungi are the main N providers in all obtained from fungi compensates for the shade. Due to mycorrhizal plants, the very different 15 N abundance in ongoing succession, the canopy of such stands tends to close mixo- and mycoheterotrophic plants, as compared to other as the forest develops, and shade increases. Mixotrophic autotrophic mycorrhizal plants ( Figure 1 b) [26] , suggests populations therefore undergo continuous selective pres- that they receive N in a different form or via a different sure for more light-independent C supply, which might pathway. Ammonium, but also some amino acids, usually lead to selection for mycoheterotrophy. A similar claim was transfers N at the mycorrhizal interface to autotrophic made for mycoheterotrophs in the Burmanniaceae that mycorrhizal plants [1,27] . Could some organic molecules evolved and diversified during the Eocene (between 56 simultaneously provide N and C to mixotrophs? Indeed, N and 34 myrs ago): at that time, high global temperatures contents of mixotrophs are often high [2,4] , but respiration enhanced expansion of tropical forests, and Burmannia- alone can account for this by increasing the N:C ratio as ceae’s autotrophic ancestors, growing in savannas, prob- compared to the food source [5] . The lack of correlation ably adapted to forest life through mycoheterotrophy [30] . between 13 C and 15 N abundances among species Interestingly, in some mixotrophic orchids, non-chlor- (Figure 1 a), as well as for a given species among sites ophyllous variants survive over years ( Figure 2 ) [25,31] . [6] , nevertheless suggests that some C is obtained through Their 13 C abundance demonstrates that they are fully different, N-independent biochemical pathways. mycoheterotrophic [4,29] . They have no photosynthetic Original mechanism(s) that might have evolved in mix- ability and respire less than green individuals ( Figure 2 ) otrophs to recover fungal C remain open to further inves- [4] . Their phenotype is stable over years, suggestive of tigation. The study of biochemical processes of C transfer either a genetic or an environmental determinism due to has been hitherto limited by the unavailability of mixo- lasting local soil conditions [29] . Interestingly, such var- trophic laboratory models due to their slow growth rates iants remain rare and produce fewer seeds because they and the complicated biological systems they rely on dry before fruit ripening [4] . Causes for this reduced fitness (Figure 1 b). Hopefully, some pyroloids can be outplanted remain to be investigated, but non-chlorophyllous variants [28] : although outplanting will destroy mycorrhizal links, are likely to represent unique snapshots of failed tran- it might allow the design of microcosms for investigating sitions from mixotrophy to mycoheterotrophy. They are mixotrophy in controlled conditions to directly demon- ecological equivalents to mutants in genetics, that is, their strate C flow to mixotrophs and identify the underlying dysfunctions might suggest what makes mycoheterotrophy biochemical pathways. successful. Although their determinism remains unknown,

67 Author's personal copy

Opinion Trends in Plant Science Vol.14 No.2

Table 1. Becoming heterotrophic by exploitation of the living source that provides mineral nutrients: a convergent scenario for evolution of heterotrophy in plants through mixotrophic steps Steps Evolution to parasitism on other plants Evolution to mycoheterotrophy Positively selected for # 1 Free-living, autotrophic mycorrhizal plant # 2 Hemiparasitic plant tapping xylem of its host Mineral nutrition by mycorrhizal fungi shared with Improved and less costly mainly for mineral nutrition surrounding plants that also contribute to fungal nutrition mineral nutrition # 3 Hemiparasitic plant tapping xylem (mineral Mixotrophic plant deriving mineral nutrients and some Improved carbon nutrition) and phloem (partial carbon carbon from shared mycorrhizal fungi, but still nutrition, improved nutrition) of its host, but still photosynthetic photosynthetic (mycorrhizal mixotrophy; Figure 1 a) tolerance to low light (parasitic mixotrophy) # 4 Mixotrophic hemiparasitic plant with reduced Mixotrophic mycorrhizal plant with reduced Improved carbon nutrition photosynthetic abilities (smaller leaves, less photosynthetic abilities (smaller leaves, less in low light pigments, etc.); better growing under its host pigments, etc.); better growing in forests # 5 Heterotrophic, non-green plant, obtaining all its Mycoheterotrophic, non-green plant obtaining all its Improved carbon mineral and carbon supply by tapping xylem mineral and carbon supply from its mycorrhizal fungi nutrition, even in absence and phloem of its host (Box 1 ) of light they offer fascinating models for comparing the physiology fungal C as mixotrophic adults. Although this syndrome of mixo- and mycoheterotrophs within similar genetic might not characterize all mixotrophs, it is likely to backgrounds. indicate a predisposition to mixotrophy at adulthood. Strikingly, the mixo- to mycoheterotrophy shift scenario Among non-flowering plants, such as ferns and club- is reminiscent of that leading to heterotrophic parasitic mosses, the small spores of several taxa develop into plants, which obtain photosynthates from host plants via a subterranean, heterotrophic gametophytes receiving C haustorium. In two lineages at least, the Orobanchaceae from their mycorrhizal fungi [40] . Recently, an association [32] and Convolvulaceae [33] , phylogenies support the with mycorrhizal fungi that simultaneously colonize sur- hypothesis that parasitic heterotrophs repeatedly arose rounding autotrophic plants was demonstrated for subter- from photosynthetic hemiparasitic ancestors. Interest- ranean stages of the fern Botrychium [41] and the ingly, as previously mentioned, several hemiparasites clubmosses Lycopodium and Huperzia [42] . Obviously, are mixotrophic, receiving some organic C from their host adult plants in these taxa are candidate mixotrophs and [12] . This suggests a common scenario ( Table 1 ), where a C deserve further attention. flow emerges in the framework of biological interactions Finally, all investigations so far have focused on forests formerly selected by mineral needs. The resulting mixo- from temperate, Mediterranean and boreal regions, where trophy, in turn, allows emergence of heterotrophy, now mixotrophs are sometimes abundant ( Figure 1 a). Dense selected by C needs only. forests, strongly selecting for alternative C sources among understory species, cover large areas in the tropics and Undiscovered mixotrophs harbour a high biodiversity. Not surprisingly, the largest Mixotrophy evolved convergently in two plant taxa (pyr- numbers of mycoheterotrophic species, although poorly oloids and orchids), probably to enable survival in the studied so far, occur in tropical forests [7] . Mixotrophic forest understory. Can we expect discoveries of more species might also have diversified in tropical environ- mixotrophic taxa? Following the model in Table 1 , myco- ments, and isotopic analyses are excellent tools for looking heterotrophy in a family is a strong predictor for mixo- for them; in such analyses, a special focus should be placed trophic species. Indeed, mycoheterotrophy arose at least on the families mentioned above. 40 times in unrelated plant families [7] where, with detailed phylogenies at hand, one could search for mixo- The ecophysiological meaning of mixotrophy trophic species. Autotrophic and mycoheterotrophic Important questions remain open on the biology of C species co-occur in many families that are excellent can- transfer. First, the cost for surrounding autotrophic plants didates for encompassing some mixotrophic species: Gen- and fungi is unknown. We do not know whether we face a tianaceae [34] , Dioscoreales [35] , Polygalaceae [36] , parasitism or a somehow balanced exchange, which could Iridaceae [37] , Pandanales [38] and Petrosaviaceae [39] . occur, for example, if mixotrophs compensate by providing Of course, one can imagine that mixotrophic ancestors of some vitamins or a ‘shelter’ (physical or chemical protec- some mycoheterotrophic lineages did not leave any extant tion) for the fungus. Second, environmental factors influ- mixotrophic descendants. At the same time, we might also encing the level of use of fungal C by mixotrophic plants are discover mixotrophic clades that never shifted to mycohe- also unclear. Many ecological or microcosm studies con- terotrophy. ducted on mixotrophic algae revealed that at least two Another feature shared by mixotrophic pyroloids and factors drive the level of heterotrophy in mixotrophic orchids might also predict mixotrophy in other plant planktonic species [43] : light deficiency favours phagocy- families. They produce a very large number of small, tosis as a source of C, whereas low nutrient levels favour sub-millimetric seeds that are devoid of reserves. Germi- phagocytosis as a source of N and P. The first surveys from nation of the undifferentiated embryo requires coloniza- orchid data suggested that light level was inversely corre- tion by the future mycorrhizal fungus for its complete lated with dependence on fungal C [2,29] , a trend that was nutrition [1] . Until expansion of the first green leaves, also observed in labelling experiments reporting C trans- subterranean seedlings are transiently mycohetero- fers among plants linked by common mycorrhizal fungi trophic, probably using the same mechanism to recover [16,17] . But this intuitively expected correlation, where

68 Author's personal copy

Opinion Trends in Plant Science Vol.14 No.2 fungal C compensates for lower photosynthesis, was not 5 Tedersoo, L. et al. (2007) Parallel evolutionary paths to observed over a wide geographical range for pyroloids [6] . mycoheterotrophy in understorey Ericaceae and Orchidaceae: ecological evidence for mixotrophy in Pyroleae. Oecologia 151, 206 – Just as for algae, plant mixotrophy might not always be 217 based on C needs, but sometimes arises as a ‘side-product’ 6 Zimmer, K. et al. (2007) Wide geographical and ecological distribution of the N and P nutrition. And this might explain the of nitrogen and carbon gains from fungi in pyroloids and monotropoids apparent discrepancies: according to the scenario in (Ericaceae) and in orchids. New Phytol. 175, 166 –175 Table 1 , organisms might (or might not) shift over evol- 7 Leake, J.R. (1994) The biology of mycoheterotrophic (‘saprophytic’) plants. New Phytol. 127, 171 –216 utionary time from a situation where C is simply hitchhik- 8 Jones, R.I. (2000) Mixotrophy in planktonic protists: an overview. ing with organic N and P to a situation where C itself is the Freshw. Biol. 45, 219 –226 main nutrient recovered. At the first step, mineral avail- 9 Heifetz, P.B. et al. (2000) Effects of acetate on facultative autotrophy in ability determines the use of fungal C, whereas at the Chlamydomonas reinhardtii assessed by photosynthetic second step, C availability (and thus light) becomes the measurements and stable isotope analyses. Plant Physiol. 122, 1439 –1445 driving factor. We might expect that, for some mixotrophic 10 Raven, J.A. (1997) Phagotrophy in photrophs. Limnol. Oceanogr. 42, species or families, soil mineral availability regulates 198 –205 dependence on fungal C. Similarly, carnivorous plants vary 11 Zubkov, M.V. and Tarran, G.A. (2008) High bacterivory by the their production of trapping devices in response to N and P smallest phytoplankton in the North Atlantic Ocean. Nature 455, 224 –226 availability [44] . Comparative analysis of several plant 12 Press, M.C. and Graves, J.D. (1995) Parasitic Plants. Chapmann et families and sites will allow testing of this hypothesis in Hall various phylogenetic backgrounds. 13 Bannister, P. and Strong, G.L. (2001) Carbon and nitrogen isotope Mixotrophy based on mycorrhizal fungi is an exciting, ratios, nitrogen content and heterotrophy in New Zealand mistletoes. – newly discovered strategy that lies inbetween auto- and Oecologia 126, 10 20 14 Adams, M.A. and Grierson, P.F. (2001) Stable isotopes at natural heterotrophy. A main goal will be to fully account for the C abundance in terrestrial plant ecology and ecophysiology: an update. budgets of mixotrophic plants over the full growth season. Plant Biol. 3, 299 –310 Another pending question is the link with experimental 15 Adamec, L. (1997) Mineral nutrition of carnivorous plants: a review. approaches demonstrating short-term C transfers by C Bot. Rev. 63, 273 –299 labelling between plants mutually sharing mycorrhizal 16 Simard, S.W. et al. (1997) Net transfer of carbon between ectomycorrhizal tree species in the field. Nature 388, 579 –582 fungi [16,19] : do these C transfers represent more than 17 Simard, S.W. and Durall, D.M. (2004) Mycorrhizal networks: a review instantaneous transfers, perhaps reversible during the of their extent, function, and importance. Can. J. Bot. 82, 1140 –1165 growing season [45] , or do they represent a significant 18 Selosse, M-A. et al. (2006) Mycorrhizal networks: des liaisons and net contribution to the C budget of receiving plants, dangereuses ? Trends Ecol. Evol. 11, 621 –628 which would therefore be mixotrophic? 19 Philip, L.J. and Simard, S.W. (2008) Minimum pulses of stable and radioactive carbon isotopes to detect belowground carbon transfer Mixotrophy emphasizes the importance of mycorrhizal between plants. Plant Soil 308, 23 –35 fungi and the sharing of common fungi with neighbours 20 Dawson, T.E. et al. (2002) Stable isotopes in plant ecology. Annu. Rev. (Figure 1 b) in plant physiology. Its discovery raises many Ecol. Syst. 33, 507 –559 questions. Is mixotrophy common in ecosystems on a global 21 Zimmer, K. et al. (2008) The ectomycorrhizal specialist orchid Corallorhiza trifida is a partial myco-heterotroph. New Phytol. 178, scale? Did it often evolve into mycoheterotrophy, and if so, 395 –400 how? What ecological factors determine the dependence on 22 Girlanda, M. et al. (2006) Inefficient photosynthesis in the fungal C? What are the cellular mechanisms whereby C is Mediterranean orchid Limodorum abortivum (L.) Swartz is mirrored transferred to receiving plants? Beyond in situ analyses of by specific association to ectomycorrhizal Russulaceae. Mol. Ecol. 15, – stable isotope contents and mycorrhizal fungal associates, 491 504 23 Nilsson, R.H. et al. (2008) Intraspecific ITS variability in the Kingdom easily tractable and relevant models, such as microcosm Fungi as expressed in the international sequence databases and its cultures, might enable broader experimental approaches implications for molecular species identification. Evol. Bioinformatics in the near future. 4, 193 –201 24 Vincenot, L. et al. (2008) Fungal associates of Pyrola rotundifolia , a Acknowledgements mixotrophic Ericaceae, from two Estonian forests. Mycorrhiza , DOI: M-A.S. is funded by the CNRS, the Agence National de la Recherche and 10.1007/s00572-008-0199-9 the French Orchid Society (SFO). The authors thank Doyle McKey, 25 Selosse, M-A. et al. (2004) Chlorophyllous and achlorophyllous Vincent Merckx, Jakub Tesitel, Bernd Zeller and two anonymous specimens of Epipactis microphylla (Neottieae, Orchidaceae) are reviewers for insightful comments and are grateful to P. Pernot for the associated with ectomycorrhizal septomycetes, including truffles. use of Figure 2 b. Microb. Ecol. 47, 416 –426 26 Trudell, S.A. et al. (2003) Nitrogen and carbon stable isotope References abundances support the myco-heterotrophic nature and host- – 1 Smith, S.E. and Read, D.J. (2008) Mycorrhizal Symbiosis. (3rd edn), specificity of certain achlorophyllous plants. New Phytol. 160, 391 Academic Press 401 2 Gebauer, G. and Meyer, M. (2003) 15 N and 13 C natural abundance of 27 Chalot, M. et al. (2006) Ammonia: a candidate for nitrogen transfer at – autotrophic and mycoheterotrophic orchids provides insight into the mycorrhizal interface. Trends Plant Sci. 11, 263 266 nitrogen and carbon gain from fungal association. New Phytol. 160, 28 Hunt, R. and Hope-Simpson, J.F. (1990) Growth of Pyrola rotundifolia – 209 –223 ssp. maritima in relation to shade. New Phytol. 114, 129 137 3 Bidartondo, M.I. et al. (2004) Changing partners in the dark: isotopic 29 Abadie, J-C. et al. (2006) Cephalantera longifolia (Neottiae, and molecular evidence of ectomycorrhizal liaisons between forest Orchidaceae) is mixotrophic: a comparative study between green – orchids and trees. Proc. Biol. Sci. 271, 1799 –1806 and nonphotosynthetic individuals. Can. J. Bot. 84, 1462 1477 4 Julou, T. et al. (2005) Mixotrophy in orchids: insights from a 30 Merckx, V. et al. (2008) Diversification of myco-heterotrophic comparative study of green individuals and non-photosynthetic angiosperms: evidence from Burmanniaceae. BMC Evol. Biol. 8, mutants of Cephalanthera damasonium . New Phytol. 166, 639 –653 178

69 Author's personal copy

Opinion Trends in Plant Science Vol.14 No.2

31 Salmia, A. (1989) General morphology and anatomy of chlorophyll-free 41 Winther, J.L. and Friedman, W.E. (2007) Arbuscular mycorrhizal and green forms of Epipactis helleborine (Orchidaceae). Ann. Bot. Fenn. symbionts in Botrychium (Ophioglossaceae). Am. J. Bot. 94, 1248 – 26, 95 –105 1255 32 Bennett, J.R. and Mathews, S. (2006) Phylogeny of the parasitic plant 42 Winther, J.L. and Friedman, W.E. (2007) Arbuscular mycorrhizal family Orobanchaceae inferred from phytochrome A. Am. J. Bot. 93, associations in Lycopodiaceae. New Phytol. 177, 790 –801 1039 –1051 43 Katechakis, A. et al. (2005) Mixotrophic versus photoautotrophic 33 McNeal, J.R. et al. (2007) Systematics and plastid genome evolution of specialist algae as food for zooplankton: the light:nutrient the cryptically photosynthetic parasitic plant genus Cuscuta hypothesis might not hold for mixotrophs. Limnol. Oceanogr. 50, (Convolvulaceae). BMC Biol. 5, 55 1290 –1299 34 Struwe, L. et al. (2002) Systematics, character evolution, and 44 Ellison, A.M. and Gotelli, N.J. (2002) Nitrogen availability alters the biogeography of Gentianaceae, including a new tribal and subtribal expression of carnivory in the northern pitcher plant, Sarracenia classification. In Gentianaceae – Systematics and Natural History purpurea . Proc. Natl. Acad. Sci. U. S. A. 99, 4409 –4412 (Struwe, L. and Albert, V.A., eds), pp. 21 –309, Cambridge 45 Lerat, S. et al. (2002) C-14 transfer between the spring ephemeral University Press Erythronium americanum and sugar maple saplings via arbuscular 35 Merckx, V. et al. (2006) Phylogeny and evolution of Burmaniaceae mycorrhizal fungi in natural stands. Oecologia 132, 181 –187 (Dioscoreales) based on nuclear and mitochondrial data. Am. J. Bot. 93, 46 Leake, J.R. (2004) Myco-heterotroph/epiparasitic plant interactions 1684 –1698 with ectomycorrhizal and arbuscular mycorrhizal fungi. Curr. Opin. 36 Imhof, S. (2008) Specialized mycorrhizal colonization pattern in Plant. Biol. 7, 422 –428 achlorophyllous Epirixanthes spp. (Polygalaceae). Plant Biol. 9, 786 – 47 Yagame, T. et al. (2007) Developmental process of an achlorophyllous 792 orchid, Epipogium roseum (D. Don) Lindl. from seed germination to 37 Reeves, P. et al. (2001) Molecular systematics of Iridaceae: evidence flowering under symbiotic cultivation with a mycorrhizal fungus. J. from four plastid DNA regions. Am. J. Bot. 88, 2074 –2087 Plant Res. 120, 229 –236 38 Rudall, P.J. and Bateman, R. (2006) Morphological phylogenetic 48 McKendrick, S.L. et al. (2000) Symbiotic germination and development analysis of Pandanales: testing contrasting hypotheses of floral of myco-heterotrophic plants in nature: transfer of carbon from evolution. Syst. Bot. 31, 223 –238 ectomycorrhizal Salix repens and Betula pendula to the orchid 39 Cameron, K.M. et al. (2003) Recircumscription of the Corallorhiza trifida Chaˆtel through shared hyphal connections. New monocotyledonous family Petrosaviaceae to include Japonolirion . Phytol. 145, 539 –548 Brittonia 55, 214 –225 49 Wickett, N.J. et al. (2008) Functional gene losses occur with minimal 40 Boullard, R. (1979) Conside´rations sur la symbiose fongique chez les size reduction in the plastid genome of the parasitic liverwort Aneura Pte´ridophytes. Syllogeus 19, 22 –23 mirabilis . Mol. Biol. Evol. 25, 393 –401

Plant Science Conferences in 2009

Protein Complexes in Plant Signalling and Development 25 –27 June 2009 University of Glasgow, UK http://www.psrg.org.uk/events/Phoenix.htm

20th International Conference on Arabidopsis Research 30 June –4 July 2009 Edinburgh, Scotland http://arabidopsis2009.com/

Plant Genomics and Beyond 5–8 July 2009 Evry, France http://www.plant-genomics.org/index.htm

70 79.(1*= 79.(1*= = = _= 8 ` = _= `= _= `= _= `= _= `= _8_= `=_= _= 5&9.&1=7*5&79.9.43=&3)=,*3*9.(=7*1&9.438-.5=4+=,7**3=&3)=3438 5-4948>39-*9.(=.3).;.):&18=.3=2.=*)=545:1&9.438=4+=*5-&1&39-*7& =47(-.)8 _== :'2.99*)=94= 1&39=.414,> _= = *8= (&8= )&1'.3.82*= (-*?= _= )&2&843.:2= *9= _= 143,.+41.& = 8439= S9433&398= (&7= (*8= .3).;.):8= )S54:7;:8= )*= (-14745->11*= 8:7;.;*39= A= 1B,*= &):19*_= *:7= 3:97.9.43= 2>(4-S9S749745-*= &= )S/A= S9S= S1:(.)S*= 2&.8= (*= 5-S342T3*= 7*89*= 9*.39S= )*= 2>89T7*`= 349&22*39= 6:&39= A= 843= )S9*72.3.82*_= 3= 3*= 8&.9= 5&8= 8.= (*= 5-S342T3*= *89= )S9*72.3S= ,S3S9.6:*2*39= *9= .1= +&:)7&.9= 54:;4.7= (74.8*7= *9= +&.7*= ,*72*7= )*8= ,7&.3*8= .88:*8= )*= (*8= (74.8*2*398= *3= (43).9.438= (43971S*8= 54:7= 54:;4.7= 1*= 8&;4.7_= 3*= &:97*= +&O43= 8*7&.9= )*= 97&3851&39*7=)*8=5.*)8=&):19*8=)&38=:3=&:97*=*3;.7433*2*39_=&1-*:7*:8*2*39`=&:(:3*=)*= (*8= )*:== *=5S7.*3(*8= 3&= S9S= 7S&1.8&'1*`= 1&= 57*2.T7*= 5&7= 2&36:*= )*= 9*258`= 5:.86:*= 1*8= (*5-&1&39-T7*8=&99*.,3*39=1S9&9=+*:.11S=&:='4:9=)*=/=&38`=*9=1&=8*(43)*=5&7=2&36:*=)*=24>*38= 9*(-3.6:*8`= 5:.86:.1= +&:)7&.9= 54:;4.7= 97&3851&39*7= 94:9= *3= 3*= '7.8&39= 5&8= 1*8= &884(.&9.438= 2>(47-.?.*33*8=)439=*11*8=)S5*3)*39=84.9=)*8=(-&25.,3438=*9=)*8=&7'7*8`=84:;*39=B,S8=4:= *3=5*72*99&39=1*:7=7*57.8*=7&5.)*=54:7=S;.9*7=1&=2479=)*8=.3).;.):8_= *99*=6:*89.43=7*89*=)43(=*3=8:85*3)=2&.8=(*79&.3*8=7S5438*8=.3).7*(9*8=439=5:=U97*= 974:;S*8=*3=S9:).&39=349&22*39=1&=)>3&2.6:*=9*2547*11*=):=5-S342T3*=a=1*8=5.*)8=&1'.348= 1*=8439=94:9=&:=143,=)*=1&=8&.843=;S,S9&9.;*=*9=):3*=&33S*=8:7=1&:97*`=(*=6:.=1&.88*=8:5548*7= 4:= '.*3= :3= )S9*72.3.82*= ,S3S9.6:*`= 4:= '.*3= 6:*= 1*= 2.(74(1.2&9= 6:.= 1*= )S9*72.3*= *89= *=97U2*2*39=89&'1*=*9=3*=)S5*3)=5&8=)*8=(43).9.438=1:2.3*:8*8=*9=7*897*.39=14(&1*2*39=(&7= 43=5*:9=974:;*7=:3=5.*)=;*79*=*9=:3=5.*)=&1'.348=A=24.38=)*=,=(2=)*=).89&3(*_= &:97*= 6:*89.43= .25479&39*= *3= 8:85*3)`= (*89= )*= 8&;4.7= (*= 6:*= 7*57S8*39*39= (*8= .3).;.):8=&1'.348=a=)*8=.3).;.):8=)*= _=)&2&843.:2 =5&72.=)&:97*8`=:3*=545:1&9.43=A=5&79=6:.= 8.3).;.):&1.8*=4:=:3*=&:97*=*85T(*='*&:(4:5=51:8=7&7*==4:7=(4257*3)7*=1S;41:9.43=)*=(*= 5-S342T3*`= *9= 349&22*39= 843= 1.*3= &;*(= 1*8= *85T(*8= (4251T9*2*39= 2>(4-S9S749745-*= (422*= _=*=.,:& =4:= _=&:89.3&* `=43=5*:9=8*=)*2&3)*7=8.=1*8=.3).;.):8=&1'.348=8*=7*574):.8*39= &;*(=1*8=.3).;.):8=;*798=4:=8.1=>=&=2.8*=*3=51&(*=)*='&77.T7*8=7*574):(97.(*8_= 4:9=)&'47)`=.1=3*=.89*=5&8=)*=2*39.43=)*=545:1&9.43=*39.T7*2*39=&1'.348=)&:(:3*= *85T(*=)*=3*499.S*8`=.1=8&,.9=94:/4:78=)*=545:1&9.438=2S1&3,S*8=41=1*8=)*:==9>5*8=).3).;.):8= +1*:7.88*39= *3= 2U2*= 9*258= *9= 1*8= (74.8*2*398= &79.+.(.*18= 24397*39= 6:*= )*8= ,7&.3*8= +*79.1*8= 5*:;*39=U97*=574):.9*8=A=1.88:=)*=(*:=8(.= (+_ &79.(1*= _== *9=&79.(1*=9*39*=)*=7S543)7*=A=(*99*=6:*89.43=A=1&.)*=)*=1&=,S3S9.6:*`=&+.3=)*=8&;4.7=8.=1*8= (74.8*2*398=8439=*++*(9.+8=*9=8.1=>=&=:3*=).++S7*3(.&9.43=*397*=1*8=)*:==9>5*8_=*9=&79.(1*= 24397*=6:*=1*8=.3).;.):8=&1'.348=3*=+472*39=5&8=:3*=*85T(*=A=5&79=*39.T7*_= &:97*=5&79`=&:= 3.;*&:=14(&1`=43=34'8*7;*=5&8=)*=).++S7*3(.&9.43=*397*=1*8=)*:==9>5*8_=3=5*:9=)43(=*3= )S):.7*=4:='.*3=6:*=1&=545:1&9.43=8*=7*574):.9=*3=7*1&9.;*=5&32.=.*`=4:='.*3=6:.1=>=&=)*= 2:19.51*8=&55&7.9.438=)&1'.348=8&38=(439*=9*=,S3S9.6:*=57S).8548S`=4:='.*3=6:*=1&1'.3.82*= )S5*3)=):3=2.(74(1.2&9_= *8=(74.8*2*398=(43971S8=24397*39=6:.1=3>=&=5&8=)*=(429=A= *++*(9:*7=)*=1&:94,&2.*`=*9= _=)&2&843.:2= *89=(438.)S7S*=(422*=51:99=&:94,&2*_= *=51:8`= 1*8=4'8*7;&9.438=)*=,>3489T2*=)*= _=)&2&843.:2 =&:=242*39=)*=1&=+147&.843=24397*39=6:*= )*8= 9:'*8= 5411.3.6:*8= >= ,*72*39= 5&7+4.8= 2U2*= &;&39= 14:;*79:7*= )*= 1&= +1*:7`= (*= 6:.= +&(.1.9*7&.9= 1&:94,&2.*= @_= @*7?*,4;&`= (422:3.(&9.43= 5*78433*11*_= *5*3)&39`= 1*8= -&25*8= +147&1*8= *3+*72S*8= 84:8= 8&(= 3439= 5&8= 94:/4:78= 57S8*39S= )*= +7:.98`= *9= 1*8= 5411.3.*8= )*8= *5-&1&39-*7& =8439=54:)7*:8*8`=&:947.8&39=A=1&=+4.8=:3*=).85*78.43=):=5411*3=14(&1*=*9=5&7=1*= 97&385479=)*8=5411.3.*8 =;.&= :3=5411.3.8&9*:7`=.1=*89=)43(=5488.'1*=6:*=1*=7S,.2*=84.9=2.=9*=(-*?= _=)&2&843.:2_= _=143,.+41.& =6:&39=A=*11*=*89=51:99=&114,&2*_= *=9*18=+1:==)*=,T3*8=5*:;*39= *=51.6:*7=1&'8*3(*=)*=).++S7*3(*=2475-414,.6:*=*397*=1*8=)*:==9>5*8=).3).;.):8_= &38=1&=2*8:7*=41=1*8=&1'.348=8439=7&7*8=*9=439=:3*=;&1*:7=8S1*(9.;*=5&7+4.8=97T8='&88*= (425&7S*=&:==.3).;.):8=;*798= (+ _= &79.(1*= `=.18=8*2'1*39=(4397*88S1*(9.433S8=*9=43=54:77&.9= 8&99*3)7*= A= ;4.7= :3*= ).++S7*3(.&9.43= ,S3S9.6:*= 2&76:&39= 1&= 8S1*(9.43_= &'8*3(*= )*= ).++S7*3(.&9.43=,S3S9.6:*=24397*=6:*=(*8=S;T3*2*398=)&55&7.9.43=8439=8&38=)4:9*=7S(*398`=*9= 2:19.51*8_= 3= (*1&`= 1&55&7.9.43= )*8= &1'.348= 3*= 8*7&.9= 5&8= ).++S7*39*= *3= +7S6:*3(*= )*= 1&55&7.9.43=(-*?=)&:97*8=*85T(*8=)*=147)7*=)*=+=(-*?= 7&'.)458.8=9-&1.&3& _=*5*3)&39`= )&38= 1*8= 545:1&9.438= A= &1'.348`= 1*:7= +7S6:*3(*= 5*:9= &99*.3)7*= 0*= = (+ _= &79.(1*= `= *9= 1&:,2*39&9.43=)*=(*99*=+7S6:*3(*=8*7&.9=):*=343=5&8=A=:3=51:8=+479=9&:==)&55&7.9.43`=2&.8=A= 1&=).85*78.43=)*=(*=(&7&(9T7*_= &114,&2.*=54:77&.9=5*72*997*=1*8=(74.8*2*398=&1'.348===;*798`= *9= '.*3= 6:*= 1*8= &1'.348= 3*= 8:7;.;*39= 5&8= 94:/4:78`= 1*= (&7&(9T7*= &1'.348= 54:77&.9= 1:.= U97*= 2&.39*3:=84:8=+472*=):3=(7>594541>2475-.82*=&11T1*8=7S(S88.+8=(&(-S8=(-*?=1*8=.3).;.):8= -S9S74?>,49*8_= *=7S,.2*=51:8=&:94,&2*=)*8= _=)&2&843.:2 =(43):.7&.9=&1478=A=1&=+7S6:*3(*= 51:8=S1*;S*=)*8=&1'.348=(-*?=(*99*=*85T(*=(425&7S=A= _=143,.+41.& _= .=1*8=S;T3*2*398=8439=2:19.51*8`=1*8=(&:8*8=)*=1&1'.3.82*=54:77&.*39=1U97*=&:88.`=*9= (*1&= 54:77&.9= *=51.6:*7= 1&= ;&7.&'.1.9S= )*8= &1'.348= )*8= ).++S7*39*8= *85T(*8= ) 5.5&(9.8`= 6:.= 57S8*39*39= ).;*78*8= ;&7.S9S8= ->54(-742*8`= &;*(= )*8= (43(*397&9.438= *3= (-14745->11*8= 4:= *3= &39-4(>&3*8=;&7.&'1*8_= &=6:*89.43=):=)S9*72.3.82*=3*89=94:/4:78=5&8=7S841:*=2&.8=(*8=)433S*8=5*72*99*39= )S(&79*7=1*8=->549-T8*8= ).841*2*39=7*574):(9*:7=*9=/:89.+.*39= 1&= (425&7&.843= )*= 1&= ;&1*:7= 8S1*(9.;*=)*8=)*:==9>5*8=).3).;.):8=54:7=(4257*3)7*=1&=8*1*(9.43=4:=(4397*88S1*(9.43=)*=9*18= &1'.348_= Plant Biology ISSN 1435-8603

RESEARCH PAPER Spatial repartition and genetic relationship of green and albino individuals in mixed populations of Cephalanthera orchids V. Tranchida-Lombardo1,2, M. Roy3, E. Bugot3, G. Santoro4,U¨ .Pu¨ ttsepp5, M. Selosse3 & S. Cozzolino2

1 Sezione di Biologia ed Ecologia Vegetale D.A.C.P.A., Universita` di Catania, Italy 2 Dipartimento di Biologia Strutturale e Funzionale, Universita` ‘‘Federico II’’, complesso MSA, Napoli, Italy 3 Centre d’Ecologie Fonctionnelle et Evolutive, CNRS, UMR 5175, Equipe Interactions Biotiques, Montpellier ce´ dex, France 4 Via della Repubblica 37, Bovino (FG), Italy 5 Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Tartu, Estonia

Keywords ABSTRACT Albinos; amplified fragment length polymorphism; mixotrophy; Several green orchids of the Neottieae tribe acquire organic carbon both mycoheterotrophy; Neottieae; photosynthesis; from their mycorrhizal fungi and from photosynthesis. This strategy may spatial repartition. represent an intermediate evolutionary step towards mycoheterotrophy of some non-photosynthetic (albino) orchids. Mixed populations of green and Correspondence albino individuals possibly represent a transient evolutionary stage offering S. Cozzolino, Dipartimento di Biologia opportunities to understand the evolution of mycoheterotrophy. In order to strutturale e funzionale, Universita` ‘‘Federico understand the emergence of albinos, we investigated patterns of spatial and II’’, complesso MSA, 80126 Napoli, Italy. genetic relationships among green and albino individuals in three mixed E-mail: [email protected] populations of Cephalanthera damasonium and one of C. longifolia using spatial repartition and Amplified fragment length polymorphism (AFLP) Editor markers. Two of these populations were monitored over two consecutive M. van Kleunen flowering seasons. In spatial repartition analyses, albino individuals did not aggregate more than green individuals. Genetic analyses revealed that, in all Received: 23 January 2009; Accepted: 14 July sampled populations, albino individuals did not represent a unique lineage, 2009 and that albinos were often closer related to green individuals than to other albinos from the same population. Genetic and spatial comparison of genets doi:10.1111/j.1438-8677.2009.00252.x from the 2-year monitoring revealed that: (i) albinos had lower survival than green individuals; (ii) accordingly, albinos detected in the first year did not correspond to the those sampled in the second year; and (iii) with one possible exception, all examined albinos did not belong to any green genet from the same and ⁄ or from the previous year, and vice versa. Our results support a scenario of repeated insurgence of the albino phenotypes within the populations, but unsuccessful transition between the two contrasting phenotypes. Future studies should try to unravel the genetic and ecological basis of the two phenotypes.

indirectly, with help of a fungal partner (Leake 1994). In INTRODUCTION the latter case, so-called ‘mycoheterotrophic’ plants have Among plants, achlorophyllous species represent fascinat- developed a dependence upon adjacent autotrophic plants ing examples of nutritional adaptation because, by lacking from which they gain carbon indirectly by the way of photosynthetic pigments, they behave as heterotrophs and shared mycorrhizal fungal partners (Leake 1994, 2004). explore alternative strategies to acquire organic carbon This peculiar capacity of obtaining carbon heterotrophi- (Graves 1995; Selosse & Roy 2009). To fulﬁl their carbon cally is particularly relevant in orchids, which comprise requirements, some plant taxa assimilate organic matter approximately 35% of the fully heterotrophic angiosperms directly from a photosynthetic source (parasitic plants) or (Waterman & Bidartondo 2008). Among orchids, the

Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands 1 Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino Patterns of albino and green Cephalanthera spp high frequency of mycoheterotrophs is a convergent evolution that occurred more than 20 times (Molvray et al. 2000) and probably derived from a peculiarity of their seed germination and embryo development. In fact, one of the most distinctive orchid characteristics is that their minute seeds contain only minimal reserves of nutrients, and depend upon fungi for the resources necessary for germination and early seedling growth (Rasmussen 1995; Smith & Read 2008). Mycoheterotrophic orchid species retain this fungus-dependent nutrition at the adult stage and remain non-photosynthetic over their lifetime (Tay- lor et al. 2002). Mycoheterotrophic orchids have been intensively investigated in the last decade and, more recently, interest has shifted to the intermediate condition in which fungi partly subsidise the nutrition of photosynthetic orchids. These plants can use fungal C in addition to their photosynthesis, a strategy called partial mycoheterotrophy or ‘mixotrophy’ (Selosse & Roy 2009). Although they are demonstrated to retain chlorophyll and photosynthesis (Julou et al. 2005; Girlanda et al. 2006), mixotrophic orchids have an isotopic composition (13C and 15N) more or less close to that of mycoheterotrophic plants due to Fig. 1. Albino and green C. damasonium individuals from Bovino, the use of fungal organic matter (Gebauer & Meyer 2003; Southern Italy. Julou et al. 2005; Abadie et al. 2006). This occurs in a number of forest-dwelling species, especially orchids that green and albino phenotype. The phenotype does not were traditionally considered as obligate autotrophs change in Cephalanthera spp. populations for up to (Gebauer & Meyer 2003; Bidartondo et al. 2004; Selosse 12 years (Julou et al. 2005; Abadie et al. 2006), which et al. 2004; Julou et al. 2005). From an evolutionary point could reflect genetic, but also micro-environmental deter- of view, mixotrophy is viewed as a predisposition to the minism of the phenotype. evolution of mycoheterotrophy (Abadie et al. 2006; Selos- Thus, spatial and genetic investigations may contribute se & Roy 2009), as it occurs in several orchid taxa where to understanding the origin(s) of albino genotypes in mixed mycoheterotrophy secondarily evolved. Among those, the populations and their genetic correlation with other green tribe Neottieae revealed to be especially pre-adapted to and albino individuals. In particular, it would be interesting mycoheterotrophy by virtue of the association of green to gather information on the following questions: species with ectomycorrhizal fungi that link them to Do albino individuals represent a unique lineage within nearby trees (Bidartondo et al. 2004; Selosse et al. 2004; each mixed population? Are the albino individuals more Julou et al. 2005). closely related to each other than to the green individu- Investigation of those intermediate mixotrophic stages, als? particularly at population level, may provide a unique Do some albino individuals belong to the same genet as opportunity to understand the evolutionary transitions to green individuals? Are there temporal transitions of the mycoheterotrophy. In this regard, achlorophyllous (= same genet between green and albino phenotype or ‘albino’) individuals that are observed in otherwise green vice versa? species, such as Epipactis and Cephalanthera spp. (Salmia How are the albino individuals spatially distributed in 1986, 1989), are of particular interest (Fig. 1). Being populations? Are albino individuals spatially closer to themselves mycoheterotrophic (Julou et al. 2005; Abadie each other than to the green individuals? et al. 2006), these albinos may document an evolutionary To answer these specific questions, we investigated the transition from mixotrophy to mycoheterotrophy. Studies patterns of spatial and genetic relationships among green on mixed populations of albino and green individuals and albino individuals in three mixed populations of have been particularly focused on the comparison of C Cephalanthera damasonium and one of C. longifolia using and N sources, fungal specificity, photosynthetic capabil- spatial repartition and markers. ity and morphology, comparing albino and green individuals in order to detect pathways of this transition. In Epipactis microphylla as well as in Cephalanthera damaso- MATERIALS AND METHODS nium and C. longifolia, albino individuals do not differ in Plant models and sampling their community of fungal symbionts from green individuals (Selosse et al. 2004; Julou et al. 2005; Abadie et al. Cephalanthera damasonium (Miller) Druce and C. longifo- 2006), suggesting a potentially easy shift between the lia (L.) Fritsch are two closely related species belonging to

2 Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands Patterns of albino and green Cephalanthera spp Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino the tribe Neottieae (Pridgeon et al. 2008). These species hypothesis. For each value of r, a 5% confidence interval are widely distributed in the Eurasian temperate zone, was obtained after 10000 repetitions of random distribu- and mainly grow on calcareous soil (Delforge 1995). We tions. L(r) was calculated for r up to 40 m for C. damaso- focused on three natural populations of C. damasonium nium and up to 4 m for C. longifolia. These scales were and one of C. longifolia (see Table 1 for locations and chosen in agreement with the population area (Fig. S1). abbreviations). The population of C. longifolia had previ- As the C. damasonium population more widely spread ously been investigated by Abadie et al. (2006) for mycor- than the C. longifolia population, the range of variation rhizal associations and C metabolism. A subsample of of r was larger for C. damasonium than for C. longifolia. each population was examined in genetic analyses. A bivariate analysis was then performed in each popu- Healthy leaves were collected from a total of 155 individ- lation, using R 2.8.0, to test the correlation between the uals (115 green and 40 albinos) and quickly stored in sil- repartitions of green and albino individuals. The null ica gel or frozen at )80 C. For the Bov C. damasonium hypothesis in this case was that green and albino sub- population, we replicated the sampling for genetic analy- populations were independent from each other, i.e., a sis in two following years (May 2005 and 2006); here, all random distribution of the phenotype (green or albino) detected albino individuals were sampled and surround- among individuals. ing green individuals were randomly sampled on a spatial basis (see Table 1). Plant genotyping For genetic analyses, total DNA was extracted from frozen Spatial analysis or silica gel-dried leaves using a modified version of the Spatial position and phenotype of each plant was cetyltrimethyl ammonium bromide (CTAB) method recorded in July 2005 for C. longifolia at Est (preci- (Doyle & Doyle 1987). Approximately 100 mg of each sion ± 0.1 m) and in July 2007 and 2008 for C. damaso- sample were incubated in 800 ll of standard CTAB buffer nium at Mon (precision ± 0.2 m; positions were tagged and 1.8 ll beta-mercaptoethanol, and incubated at 60 C in situ in 2007 and 2008 for phenotype monitoring). A for 30 min. Subsequently, the mixture was extracted twice database was generated for C. damasonium data using with an equal volume of chloroform:isoamylic alcohol ArcGIS 9.2 (ESRI, Redland, CA, USA) to create maps of (24:1) and centrifuged at 6000 g for 10 min. DNA was iso- the populations. Minimum distances between two indi- propanol precipitated at 10000 g and 4 C for 20 min, viduals were computed with ArcGIS 9.2 and analysed washed with 70% cold ethanol (4 C) for 5 min and air using Minitab statistical software 13.31 (Minitab Inc., dried for 15–30 min. DNA pellets were resuspended in State College, PA, USA). The matrix of distances was 30 ll distilled water and quality of DNA was examined computed with ArcGIS 9.2 and its correlation with the electrophoretically on 0.8% agarose gels. matrix of genetic distances (see below) was tested using a Amplified fragment length polymorphism (AFLP) pro- Mantel test, with R 2.8.0 (http://cran.r-project.org/). cedure was performed as in Vos et al. (1995) with minor To analyse interplant distances, all data were intro- modifications, as reported in Moccia et al. (2007) and duced into R 2.8.0, and univariate spatial patterns (aggre- using fluorescent dye-labelled primers. A preliminary test gation ⁄ regularity) of plants were analysed using Ripley’s was conducted with 11 primer pairs on a sample subset K(r) function (Ripley 1977). The K(r) function is defined of 10 individuals for each population. The size of AFLP as the expected number of plants within distance r from fragments was determined with the software GENEMAP- a randomly chosen plant. Assuming complete spatial ran- PER v3.0 (Applied Biosystems, Foster City, CA, USA) and domness, K(r) = pr2 (for detailed properties of the func- amplified traces were scored as present or absent in a tion K(r), see Diggle 1983). Assuming a complete spatial binary data matrix. In order to test the reproducibility of independence, the derived function LðrÞ¼½KðrÞ=p0:5–r AFLP profiles, we replicated the DNA extraction proce- has thus an expected value of zero. We tested whether dures and amplification protocols on 22 randomly selected observed distributions were more aggregated or more reg- individuals (four out of 49 from Est, eight out of 45 from ular than random distributions expected from the null Mon, four out of 42 from Bov and six out of 19 from

Table 1. Geographic location, population abbreviation, year of sampling for spatial and genetic analyses and sample size for genetic analysis of green and albino individuals in populations of Cephalanthera longifolia and C. damsonium.

population year(s) of spatial year(s) of green species location abbreviation analysis genetic sampling individuals albinos N E

C. longifolia Estonia Est 2005 2005 44 5 5814¢44¢’2200¢41¢’ C. damasonium Southern France Mon 2007 and 2008 2006 27 18 4339¢27¢¢ 351¢53¢¢ C. damasonium Southern Italy Bov – 2005 and 2006 12 and 17 6 and 7 4114¢49¢’1521¢00¢’ C. damasonium Northern Italy Vic – 2006 15 4 4532¢13¢’1132¢25¢’ total 4 populations 115 40

Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands 3 Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino Patterns of albino and green Cephalanthera spp

Vic). Only fragments with homogeneous, strong intensity The significance of the mean differences between He were included in the data matrix; loci that did not give and I values between albino and green individuals was clear (i.e., reproducible and easily scored) signals were assessed by paired sample t-tests carried out using SPSS excluded from the analysis. Loci accumulating too many 13.0 statistical package (SPSS Inc., Chicago, IL, USA). differences (i.e., 5% between two replicates) were consid- Furthermore, using the same parameters, for the Bov ered as prone to genotyping errors and were discarded. population in which two consecutive samplings were available, we also analysed the spatio-temporal genetic diversity across 2005 and 2006 sampling using the POP- AFLP data analysis GENE version 1.31 program (Yeh et al. 1999). Cephalanthera damasonium is a self-pollinating plant, For all populations, the genetic relationships among while C. longifolia is a facultative allogamous plant (Scac- individuals was examined with a UPGMA dendrogram chi & de Angelis 1991). Consequently, a low proportion based on the Jaccard similarity calculated as J = a ⁄ (n)d), of heterozygosity is expected and analysis of dominant where a is the number of positive matches (i.e., the pres- markers like AFLP becomes less problematic (Lynch & ence of a fragment in all samples), d is the number of Milligan 1994). Each peak in the AFLP fingerprint pattern negative matches (i.e., the absence of a fragment from all was considered as a separate putative locus; all genotypes samples), and n is the total sample size, including both were scored for the presence (1) and absence (0) of poly- matches and mismatches. Dendrograms were obtained morphic AFLP fragments as dominant markers (Pomp- with the FAMD software (Schlu¨ter & Harris 2006). anon et al. 2005), and the data were considered as haploid and entered into a binary matrix. The binary matrices of AFLP phenotypes were then assembled for RESULTS statistical and genetic analyses. We calculated the total Spatial distribution and survival of green and albino number of scored fragments, number of fixed fragments individuals and percentage of polymorphic fragments. For each locus, we calculated the genotyping error rate by numbering the The Est C. longifolia population had eight albinos and allelic differences between genotypes obtained from the 237 green individuals in an area of 585 m2 in 2004 same sample after separate DNA extraction and amplifica- (Fig S1). The Mon C. damasonium population had 50 tion. The error rate per locus was then calculated as the albinos and 592 green individuals in an area of 5096 m2 allelic differences between the genotypes obtained for each in 2007, and 44 albinos and 825 green individuals in 2008 of the two replicates divided by the total number of frag- (Supporting Information). The percentage survival from ments per profile (Bonin et al. 2004). All screenings were 2007 to 2008 was 86% for green individuals and 60% for performed twice in two independent projects created with albino individuals. We also found 316 new green individ- software GENEMAPPER 3.0 (Applied Biosystems) and uals and 14 new albinos in 2008. Production of new results were compared as suggested by Bonin et al. shoots in two successive years at Mon was thus higher for (2004). green (+233) than for albino ()6) plants. No phenotype Similarity of AFLP profiles between pairs of individuals shift was observed in any individual at Mon. in each population was calculated using the similarity In the Mon C. damasonium population, the L(r) func- index (Lynch 1988, 1990) as: Sxy¼ 2nxy=ðnxþnyÞ, where tion of green individuals revealed two scales of aggrega- nx and ny are the number of fragments present only in tion at 10 and 40 m (Fig. 2a), and the L(r) function of sample x and y, respectively, and nxy is the number of albinos revealed a single scale of aggregation at 12.5 m fragments common for samples x and y. The same analy- (Fig. 2b). Bivariate analysis of the repartition of green sis was also conducted between all replicates of the 22 versus albino individuals revealed no aggregation between randomly selected individuals to obtain the mean value of the two phenotypes (Fig. 2c). For the Est C. longifolia similarity between replicates (i.e., clones) of the same population, the L(r) functions revealed aggregations for accession. both phenotypes, around 1.5 m for green individuals and When applied to predominantly self-pollinating plant 1.1 and 1.5 m for albino plants (Fig. 2d and e). The species where heterozygotes are rare, the gene diversity bivariate analysis failed to show any correlation between index should yield relatively accurate estimations also for green and albino individuals of C. longifolia (Fig. 2f). dominant markers (Lynch & Milligan 1994). To analyse However, the low number of albinos in Est limits the the within-population gene diversity between subsets of power of these analyses, as reflected in the large 5% con- albino and of green individuals,P we used: (a) the mean of fidence interval for L(r). We therefore also investigated Nei’s gene diversity ðHe ¼ 1 piÞ (Nei 1973) com- minimal spatial distances between phenotypes. In both puted as the expected heterozygosity and based on Mon and Est populations, the mean values significantly allele frequencies for each locus and for all loci, both decreased in the order: green to nearest albino > albino polymorphic and non-polymorphic (Ozbek et al. 2007); to nearest albino > albino to nearest green > green to andP (b) the mean of Shannon’s information index nearest green (Table 2). Thus, albinos were less spatially ðI ¼ pi ln piÞ (Lewontin 1972), indicating the degree aggregated than green individuals, and were closer to of marker polymorphism within populations. green than to albino individuals.

4 Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands Patterns of albino and green Cephalanthera spp Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino

abc

def

Fig. 2. Uni- and bivariate spatial patterns of individuals using the L(r) function. Assuming complete spatial independence between individuals, L(r) has an expected value of zero; the two dotted lines around zero indicate the 5% conﬁdence interval. (a–e), univariate analyses: L(r) calculated for Mon (a, green individuals and b, albino individuals) and Est (d, green individuals and e, albino individuals) populations. (c–f), bivariate analyses: L(r) calculated for Mon (c) and Est (f).

Table 2. Analysis of minimal inter-plant distance between individuals als and Methods). The percentage error per duplicate at Mon and Est. For each population, values followed by different let- sample, defined as the frequency of differences observed ters differ significantly according to an ANOVA (P < 0.001). A Tukey between two runs of the same sample (Bonin et al. 2004), post hoc test was not applied because of samples of different size. varied from 1.5% in the primer combination EcoACG- MseAGAC to 4.33% in EcoAGC-MseCAC; the mean per- average minimal distance between Mon Est centage of errors for all the primer combinations was green and nearest green individual 0.70 ± 0.87 a 0.30 ± 0.23 a 2.93% (SD ± 0.95). The mean difference between the two green and nearest albino individual 7.92 ± 5.58 b 3.03 ± 2.13 b independent scorings was 2.69% (SD ± 0.42) (Est: 3.10%; albino and nearest green individual 0.91 ± 0.74 c 0.40 ± 0.20 c Mon: 2.95%; Bov: 2.18%; Vic: 2.52%), in agreement with albino and nearest albino individual 2.35 ± 2.72 d 1.92 ± 1.51 d previous reports (below 5%; Bonin et al. 2004). The mean similarity index in AFLP profiles between pairs of replicated individuals (equated to clones) was 0.98 (ranging from 0.97 to 1.0); the mean similarity index between individual pairs (resulting from dendrograms) in the Genetic variability and marker reproducibility examined populations was 0.93 (ranging from 0.85 to Among the 11 EcoRI ⁄ MseI primer pairs tested, six highly 0.98; Fig. 3). polymorphic combinations with a large number of clear fragments were selected for the analysis. Using these six Genetic analyses AFLP primer combinations, the 155 individuals produced a varying number of fragments per combination Values of Nei’s gene diversity (He) and Shannon’s infor- (Table 3). The mean number of fragments in each primer mation index (I) of the population of C. longifolia (Est) combination ranged from 17.7 (EcoAGC-MseACAC) to were in the range of values found for the three popula- 41.0 (EcoAGG-MseAGAC) (mean 26.12, SD ± 8.52). The tions of C. damasonium (Fig. 4). Nei’s gene diversity (He) mean number of monomorphic (or non-polymorphic) and Shannon’s information index (I) did not differ signif- fragments ranged from 3.25 (EcoAGG-MseAGAC, EcoAGC- icantly between albino and green individuals within pop- MseACAC) to 7.75 (EcoACG-MseCTG) (mean 5.20, ulations (He: t = 0.38, P = 0.72. I: t = 0.10, P = 0.92) SD ± 1.71). The percentage of polymorphic fragments (Fig. 4), indicating that genetic diversity was similar for considering all primer combinations ranged from 66.4% both phenotypes. This also appeared to be consistent in (Bov population) to 83.7% (Est population; see Table 3). the Bov population across the two consecutive years 2005 The reproducibility of scoring of the AFLP patterns and 2006 (Fig. 4). was checked by analysing the genotyping error rate in The UPGMA dendrograms (Fig. 5) illustrating genetic repeated extractions and runs of 22 replicates (see Materi- similarity among individuals showed that the albino indi-

Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands 5 Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino Patterns of albino and green Cephalanthera spp

Table 3. Number of AFLP fragments and levels of polymorphism among all samples in each examined population (see Table 1 for population abbreviations).

population

Est Mon Bov Vic

no. of % poly- no. of % poly- total no. of % poly- total no. of % poly- primer total no. fixed morphic total no. fixed morphic no. of fixed morphic no. of fixed morphic combinations of bands bands bands of bands bands bands bands bands bands bands bands bands

EACG-MCCAA 40 8 80.00 20 3 85.00 26 6 77.00 21 1 95.23 EACG-MAGAC 26 5 80.76 32 4 87.50 29 5 82.75 32 8 75.00 EACG-MCTG 25 6 76.00 14 7 50.00 29 11 62.06 16 7 56.00 EACG-MCAC 22 6 72.72 18 0 100 22 16 27.27 20 4 80.00 EAGC-MACAC 14 0 100 17 3 82.35 14 5 64.28 26 7 73.00 EAGG-MAGAC 41 3 92.68 38 2 94.73 48 7 85.41 37 1 97.00 total over 6 loci 168 28 83.69 139 19 83.26 168 50 66.46 152 28 79.37

Fig. 3. Similarity index (horizontal axis) of AFLP proﬁles between rep- lication pairs of the same samples (white) and between two individuals (black) of each terminal pair in UPGMA dendrograms (as displayed in Fig. 5) for all populations. Numbers of pairs are plotted on the vertical axis. viduals formed neither a unique phyletic lineage nor a unique clone within each population. Even if, in a few circumstances, some albino individuals partially clustered together (e.g., at Mon), generally they were scattered Fig. 4. Genetic diversity within-population between albino (white col- throughout the dendrograms and, as a result, seemed umns) and green individuals (black columns) based on Nei’s gene more genetically similar to some green individuals. diversity (He) values and Shannon’s information index (I). Among paired individuals in each population, seven pairs displayed levels of similarity index ranging from 0.97 to ) 0.98 that may support either a clonal origin or a close ered. It ranged from 1 20 (most common AFLP pheno- ) kinship, given the percentage of error per duplicate type, with most frequent alleles at each locus) to 1.66 65 sample (Fig. 3). No such pairs included two albinos, and (less common AFLP phenotype, with less frequent alleles among these, a single individual pair (at Bov) encom- at each locus). The probability that a green and an albino passed an albino and a green individual (sampled in 2005 individual would share the same AFLP pattern by chance ) and 2006, respectively; similarity index: 0.97). In the Bov was 1.49 31. population, the probability that two individuals have the same AFLP pattern by chance can be calculated by taking Spatial genetic structure at Est and Mon into account the frequency of absence ⁄ presence of each AFLP band. Assuming that bands are genetically indepen- According to a Mantel test, genetic distance between indi- dent, the probability of an AFLP pattern is p = JIfi viduals (not considering their phenotype) was signiﬁ- where fi is the phenotypic frequency for AFLP band i cantly correlated with geographic distance (r = 0.149, (presence or absence of band i) in the population consid- P = 0.022) in the Mon population, as was also the case

6 Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands Patterns of albino and green Cephalanthera spp Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino

Fig. 5. UPGMA dendrograms based on Jaccard genetic distance among individuals in the four examined populations. Empty circles indicate albino individuals, and black vertical lines indicate individuals that potentially belong to the same clone.

for the subset of green individuals (r = 0.173, P = 0.018). whole population in Mon (r = 0.142, P = 0.006). Even However, the correlation was not significant for the sub- when only investigating albino to green individual dis- set of albinos in the Mon populations (r = 0.095, tance, no strong correlation was detected between genetic P = 0.228). For the Est population, the results were simi- and geographic distance in Mon (r = 0.022, P = 0.001). lar; the correlation between genetic and geographic dis- However, the opposite result was found in Est, albino tance was significant when all individuals were included and green individuals closely genetically related were also (r = 0.619, P = 0.001) and for the subset of green indi- geographically closer (r = 0.741, P < 0.0001) but the cor- viduals (r = 0.683, P = 0.001), but not for the subset of relation was not significant when considering the whole albinos (r = 0.726, P = 0.071). In all, although more sim- population (r = 0.0079, P = 0.051). Such results under- ilar genotypes tended to group together spatially in popu- line the absence of a genetic structure for both albino and lations, we did not find clear genetic aggregation on the green individuals.

Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands 7 Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino Patterns of albino and green Cephalanthera spp

DISCUSSION individuals. Because these genetically related albinos do not represent clones, their genetic proximity could be due Possible origin of albinos to reproduction, perhaps autogamously, of an albino or a Albino individuals in populations of usually green orchids green individual with a mutation in a generative cell that represent a hitherto poorly analysed phenomenon that generated progeny of closely related albino individuals. could be an intermediate step in the evolutionary emer- Since the levels of genetic diversity in the subset of albino gence of mycoheterotrophy. Previous studies failed to show individuals are equivalent to those of corresponding green any significant difference in mycorrhizal colonisation and individuals within the same mixed population, this would morphology between albino and green phenotypes in indicate that the level of selfing or outbreeding reproduc- mixed population of C. longifolia (Abadie et al. 2006) and tion in green and albino individuals are comparable and of C. damasonium (Julou et al. 2005). However, the high that non-assortative mating does not occur between the variance in mycorrhizal fungi and individual development two different phenotypes. It is, however, not known may obscure differences. Without doubt, albinos belong to whether albinos can reproduce. the same species as green individuals, based on ribosomal internal transcribed sequence identification, flower Spatial distribution morphology and crossing experiments (Me´lanie Roy & Marc-Andre´ Selosse, unpublished data). Although the There is evidence that albinos do not necessarily occur mechanism responsible for the albino phenotype is near green individuals (Julou et al. 2005). The Bov and unknown, our study shows that albino individuals do not Vic populations also suggest that some albinos occurred constitute single isolated lineages within each population. at large distances from green and other albino individuals Although parentage analysis are almost impossible to (data not shown). Our spatial analysis at Est and Mon carry out using dominant markers, the genetic relation- suggest that, although albinos occur within dense patches ship among green and albino individuals within the four of green individuals, they do not tend to be closer to examined mixed populations shows that albinos do not other orchids (albino or green) than to green orchids represent a monophyletic lineage within each population (Table 2) and no correlation was revealed between albino (Fig. 5). Also, the absence of spatial aggregation of albi- and green individual distribution in Mon and Est nos at small scales (Fig. 2) suggests that we do not face (Fig. 2). These observations are relevant, since the C is the spread of a single clone (by vegetative propagation) in conveyed to mycoheterotrophic orchids from surrounding each population. The over-dispersion of albinos in dendr- green plants through shared mycelial partners. Assuming ograms (Fig. 5) and spatial repartition (Table 2) suggest that roots donating C need to be spatially close (Selosse that the albino phenotype occurred repetitively within et al. 2002), these observations further corroborate that different progenies or different plants. This could explain the ultimate C sources are not green conspecific individu- why, in the analysed populations, albinos are more closely als (Julou et al. 2005), but more likely other fully auto- related to green individuals than to other albinos. This trophic plants, such as nearby trees, that supply C to the scenario is congruent with the fact that we recover similar shared mycorrhizal fungi. levels of genetic diversity for albinos in C. damasonium and C. longifolia, in spite of the presumed differences in Temporal trends their mating system (mostly autogamous versus mostly allogamous; Scacchi & de Angelis 1991). Most of the time, albino orchids are rare and do not The genetic basis for albinism remains unknown. Albi- invade the populations where they occur (Renner 1938; nos could be either permanent mutants, as suggested by Mairold & Weber 1950). Congruently, albinos were seen phenotype stability over the years (see below), or transitory at constantly low frequency, with slower demography phenotypic stages, in which genes involved in the photo- than green individuals. In Mon, twice as many albinos synthetic pathway can switch off depending on micro-envi- died or became dormant as compared to green individu- ronmental conditions (e.g., the amount of C resources als; albino number increased more slowly in 2008 (+28%) provided by the nearby fungal mycelia or tree roots) that than the number of green individuals (+40%). prevent greening. The results of our study are congruent In the Bov population, where an exhaustive sampling with both hypotheses. Noteworthy, under the hypothesis of of the albinos was performed over 2 years, genetic analy- environmental determinism, the spatial distribution of ses demonstrated that albinos of two different years albinos would indicate that conditions enhancing this phe- always showed divergent AFLP patterns. According to the notype are very heterogeneously distributed, occurring in efficiency of detecting clones in our AFLP approach very small patches (Supporting Information), and are very (Fig. 3), we can confidently exclude that they belonged to stable over years, so that albinos do not group together and the same genet (unfortunately, no spatial data are avail- remain phenotypically stable (Table 2; Fig. 2). able for the Bov population). Congruent with the absence The partial genetic clustering of albinos in the of shoots from the same genet over 2 years, the albinos C. damasonium Mon population (although they remain collected in two different years at Bov were also geneti- genetically different from each other; Fig. 5) may indicate cally more closely related to green individuals than to a closer relationship among some albinos than with green each other. The observation that many albinos did not

8 Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands Patterns of albino and green Cephalanthera spp Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino form a shoot in the next year may either suggest lower try to unravel the genetic and ecological basis of the two survival or that they remain dormant underground after a phenotypes. blooming season. The latter condition has been reported for green Cephalanthera sp. individuals (Rasmussen 1995; SUPPORTING INFORMATION Shefferson et al. 2003) and was described in previous monitoring of albino individuals (Abadie et al. 2006). Additional Supporting Information may be found in the As previously reported in the Est C. longifolia popula- online version of this article: tion (Abadie et al. 2006), the 2-year samplings in Bov Figure S1. Map of the green (black circles) and albino and Mon populations confirmed the stability of the (open circles) individuals of two investigated populations, albino phenotype over years. At least, the transition Mon (a, C. damasonium in 2008) and Est (b, C. longifolia between green and albino phenotypes within the same in 2004). genet is an uncommon event (no change among the 509 Please note: Wiley-Blackwell are not responsible for the green and 30 albino individuals observed in both 2007 content or functionality of any supporting materials sup- and 2008 at Mon). In the 2-year exhaustive sampling of plied by the authors. Any queries (other than missing albinos in Bov population, we only found one case that material) should be directed to the corresponding author could represent a temporal transition between pheno- for the article. types – i.e., a green individual that could be part of an albino genet sampled the year before. However, due to ACKNOWLEDGEMENTS the significant probability of obtaining a similar genotype only by chance, we cannot firmly reject the alternative The authors thank Daniele Doro, for providing detailed hypothesis that the albino and the green individual of information on the population at Vicenza. They also this pair represent two genetically closely related but dif- thank A. Croce, G. Scopece, N. Juillet, A. Palo and U. ferent genets. Moreover, given the potential existence of Pu¨ttsepp for help with field sampling, as well as S. Gior- a dormant stage and the lack of spatial data at Bov, it gio, G. Napolitano and members of staff of Naples Bota- could also be that the albino is of recent origin, and that nic Garden for support in plant collection. We also thank the green individual was not detected in 2006. With this the two reviewers and the Associate Editor, whose com- potential exception, all examined albino shoots never ments provided a significant contribution to improve- belonged to green genets from the same or previous ment of the article. Funding for this study was from the flowering year. In long-term investigations of Est (Abadie 2007 Programma di Rilevante Interesse Nazionale (PRIN) et al. 2006), as in our observations, the lack of interme- to S. C. and the Centre National de la Recherche Scientif- diate phenotypes further corroborated the stability of the ique (CNRS) and the Socie´te´ Française d’Orchidophilie albino phenotype. A different situation was observed for (SFO) to M.-A. S. Epipactis helleborine, where intermediate (yellowish-green or variegated) phenotypes occur (Salmia 1989), and for REFERENCES E. neerlandensis where albinos sometimes reverse to yellowish-green over a growing season (Lemagnen & Selos- Abadie J.C., Pu¨ttsepp U¨ ., Gebauer G., Faccio A., Bonfante P., se, unpublished data). Selosse M.A. (2006) Cephalanthera longifolia (Neottieae, Orchidaceae) is mixotrophic: a comparative study between CONCLUSIONS green and non-photosynthetic individuals. Canadian Journal of Botany, 84, 1462–1477. In the investigated Cephalanthera populations, we found Bidartondo M.I., Burghardt B., Gebauer G., Bruns T.D., Read a limited spatial structure (with similar genotypes group- D.J. (2004) Changing partners in the dark: isotopic and ing together) and high level of genetic diversity within molecular evidence of ectomycorrhizal liaisons between for- and among populations. Albinos were as genetically est orchids and trees. Proceedings of the Royal Society of Lon- diverse as green individuals, and did not show any signifi- don. Series B, Containing papers of a Biological character. cant spatial structure. However, the low number of albi- Royal Society (Great Britain), 271, 1799–1806. nos limited our analyses in terms of statistical power. Bonin A., Bellamain E., Bronken Eidesen P., Pompanon F., Two-year monitoring suggested lower survival or higher Brochmann C., Taberlet P. (2004) How to track and assess shift to dormancy of albinos as compared to green individuals. In our study, within the limits of our sampling, genotyping errors in population genetics studies. Molecular albinos do not represent a unique lineage within each Ecology, 13, 3261–3273. examined populations and do not tend to cluster together Delforge P. (1995) Orchids of Britain & Europe. Harper Collins, either spatially or genetically. In conclusion, our results London, UK. support a scenario of a repeated insurgence of the albino Diggle P.J. (1983) Statistical analysis of spatial point patterns. phenotypes within the populations, but unsuccessful tran- Academic Press, New York, USA. sition between the two contrasting phenotypes. Although Doyle J.J., Doyle J.L. (1987) A rapid procedure for DNA puri- experimental work on mixotrophic orchids is challenging fication from small quantifies of fresh leaf tissue. Phytochem- (Sadovsky 1965; Rasmussen 1995), future studies should ical Bulletin, 19, 11–15.

Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands 9 Tranchida-Lombardo, Roy, Bugot, Santoro, Pu¨ ttsepp, Selosse & Cozzolino Patterns of albino and green Cephalanthera spp

Gebauer G., Meyer M. (2003) 15N and 13C natural abundance Rasmussen H.N. (1995) Terrestrial orchids – from seed to myco- of autotrophic and myco-heterotrophic orchids provides trophic plant. Cambridge University Press, Cambridge, UK. insight into nitrogen and carbon gain from fungal associa- Renner O. (1938) U¨ ber blasse, saprophytische Cephalanthera tion. New Phytologist, 160, 209–223. alba und Epipactis latifolia. Flora, 132, 225–233. Girlanda M., Selosse M.A., Cafasso D., Brilli F., Fabbian R., Ripley B.D. (1977) Modelling spatial patterns. Journal of the Ghignone S., Pinelli P., Segreto R., Loreto F., Cozzolino S., Royal Statistical Society. Series B, Statistical methodology, 39, Perotto S. (2006) Inefficient photosynthesis in the Mediter- 172–212. ranean orchid Limodorum abortivum is mirrored by specific Sadovsky O. (1965) Orchideen im eigenen Garten. BLV, Mu¨n- association to ectomycorrhizal Russulaceae. Molecular Ecol- chen, Germany. ogy, 15, 491–504. Salmia A. (1986) Chlorophyll-free and green forms of Epipactis Graves J.D. (1995) Host-plant responses to parasitism. In: helleborine (Orchidaceae) in SE Finland. Annales Botanici Press M.C., Graves J.D. (Eds), Parasitic plants. Chapman Fennici, 23, 49–57. and Hall, London, UK: pp. 206–225. Salmia A. (1989) General morphology and anatomy of chloro- Julou T., Burghardt B., Gebauer G., Berveiller D., Damesin C., phyll-free and green forms of Epipactis helleborine (Orchida- Selosse M.A. (2005) Mixotrophy in orchids: insights from a ceae). Annales Botanici Fennici, 26, 95–105. comparative study of green individuals and non-photosyn- Scacchi R.G., de Angelis R.M. (1991) Effect of the inbreeding thetic individuals of Cephalanthera damasonium. New Phy- system on the genetic structure in three Cephalanthera spp. tologist, 166, 639–653. (Orchidaceae).. Plant Systematics and Evolution, 176, 53–62. Leake J.R. (1994) The biology of myco-heterotrophic (‘sapro- Schlu¨ter P.M., Harris S.A. (2006) Analysis of multilocus finger- phytic’) plants. New Phytologist, 127, 171–216. printing data sets containing missing data. Molecular Ecology Leake J.R. (2004) Myco-heterotroph ⁄ epiparasitic plant Notes, 6, 569–572. interactions with ectomycorrhizal and arbuscular mycorr- Selosse M.-A., Roy M. (2009) Green plants eating fungi: facts hizal fungi. Current Opinion in Plant Biology, 7, 422–428. and questions about mixotrophy. Trends in Plant Science, Lewontin R.C. (1972) The apportionment of human diversity. 14, 64–70. Evolutionary Biology, 6, 381–398. Selosse M.-A., Weiss M., Jany J.-L., Tillier A. (2002) Commu- Lynch M. (1988) Estimation of relatedness by DNA finger- nities and populations of sebacinoid basidiomycetes associ- printing. Molecular Biology and Evolution, 5, 584–599. ated with the achlorophyllous orchid Neottia nidus-avis (L.) Lynch M. (1990) The similarity index and DNA fingerprinting. LCM Rich. and neighbouring tree ectomycorrhizae. Molecu- Molecular Biology and Evolution, 7, 478–484. lar Ecology, 11, 1831–1844. Lynch M., Milligan B.G. (1994) Analysis of population Selosse M.-A., Faccio A., Scappaticci G., Bonfante P. (2004) genetic-structure with RAPD markers. Molecular Ecology, 3, Chlorophyllous and achlorophyllous specimens of Epipactis 91–99. microphylla (Neottieae, Orchidaceae) are associated with Mairold F., Weber F. (1950) Notiz u¨ber Cephalanthera albinos. ectomycorrhizal septomycetes, including truffles. Microbial Protoplasma, 39, 275–277. Ecology, 47, 416–426. Moccia M.D., Widmer A., Cozzolino S. (2007) The strength of Shefferson R.P., Proper J., Beissinger S.R., Simms E.L. (2003) reproductive isolation in hybridizing food deceptive orchids. Life history trade-offs in a rare orchid: the costs of flower- Molecular Ecology, 16, 2855–2866. ing, dormancy, and sprouting. Ecology, 84, 1199–1206. Molvray M., Kores P.J., Chase M.W. (2000) Polyphyly of Smith S.E., Read D.J. (2008) Mycorrhizal Symbiosis. 3rd edn. mycoheterotrophic orchids and functional influences on floral Academic Press, London, UK. and molecular characters. In: Wilson K.L., Morrison D.A. Taylor D.L., Bruns T.D., Leake J.R., Read D.J. (2002) Mycorrhi- (Eds), Monocots: systematics and evolution. CSIRO, Mel- zal specificity and function in myco-heterotrophic plants. bourne, Australia, pp. 441–447. In: Sanders I., van der Heijden M.G.A. (Eds), Mycorrhizal Nei M. (1973) Analysis of gene diversity in subdivided popula- Ecology. vol. 157. Springer, Berlin, Germany: pp. 157. tions. Proceedings of the National Academy of Sciences USA, Vos P., Hogers R., Bleeker M., Reijans M., Van de Lee T., 70, 3321–3323. Hornes M., Fritjers A., Pot J., Peleman J., Kuiper M., Ozbek O., Millet E., Anikster Y., Arslan O., Feldman M. (2007) Zabeau M. (1995) AFLP – a new technique for DNA-finger- Spatio-temporal genetic variation in populations of wild printing. Nucleic Acids Research, 23, 4407–4414. emmer wheat, Triticum turgidum ssp. dicoccoides, as revealed Waterman R.J., Bidartondo M.I. (2008) Deception above, by AFLP analysis. Theoretical and Applied Genetics, 115, 19–26. deception below: linking pollination and mycorrhizal biol- Pompanon F., Bonin A., Bellemain E., Taberlet P. (2005) ogy of orchids. Journal of Experimental Botany, 59, 1085– Genotyping errors: causes, consequences and solutions. Nat- 1096. ure Reviews Genetics, 6, 847–859. Yeh F.C., Yang R.C., Boyle T.. (1999) POPGENE version 1.31. Pridgeon A.M., Cribb P.J., Chase M.W., Rasmussen F.N. Microsoft Windows-based freeware for population genetic anal- (2008) Genera Orchidacearum. Vol. 4: Epidendroidea (part ysis. University of Alberta and Centre for International one). Oxford University Press, Oxford, UK. Forestry Research, Alberta, Canada.

10 Plant Biology ª 2009 German Botanical Society and The Royal Botanical Society of the Netherlands 79.(1*= 79.(1*= =

= _= `=_= `= _= `= _8_=`=_= `=_=`=_= `=_8_= _= 4385-4948>39-*9.(= &1'.34= ;&7.&398= .3= *5-&1&39-*7&= )&2&843.:2 a=&=)*&)8*3)=<&>=.3= *;41:9.43=94<&7)=2>(4-*9*749745-> = 3=57*5_ = = &79.(1*= 57S(S)*39= 34:8= &= 24397S= 1-424,S3S.9S= ,S3S9.6:*= )*8= 545:1&9.438= )*= _= )&2&843.:2 =6:.=5*72*9=)*=(4257*3)7*=1&'8*3(*=)*=).++S7*3(.&9.43=2475-414,.6:*=4'8*7;S*_= *5*3)&39`= (*99*= &'8*3(*= )*= ).++S7*3(*= *89= .397.,&39*= 5:.86:*= (*8= &1'.348= 8439= 7&7*2*39= 974:;S8=)&38=1*8=545:1&9.438=*9=94:/4:78=*3=+&.'1*=+7S6:*3(*_=3=5*:9=)43(=8*=)*2&3)*7=(*= 6:.=54:77&.9=(4397*=8S1*(9.433*7=(*=9>5*`=4:='.*3=(*=6:.=54:77&.9=1*=8S1*(9.433*7=*9=(4389.9:*7= &.38.=:3*=*85T(*=2>(4-S9S749745-*`=(422*=.1=*3=*=.89*=)&38=1*=,*37*= *5-&1&39-*7& _= +.3= )*= 2.*:== (4257*3)7*= 1S;41:9.43= )*= 1&= 2>(4-S9S749745-.*= )&38= 1*= ,*37*= *5-&1&39-*7& `=)*:==545:1&9.438=2.=9*8=)*= _=)&2&843.:2 =439=S9S=S9:).S*8=5*3)&39=974.8=&38= 84:8=)*=342'7*:==&3,1*8=a=A=1&=+4.8=8:=54.39=)*=;:*=)*=1&=)>3&2.6:*=85&9.&1*`=9*2547*11*`=)*= 1&=8:7;.*=&:=(4:78=)*=1&33S*`=):=)S;*1455*2*39`=)*=1&=5->8.414,.*=)*8=+*:.11*8`=):=7S,.2*=)*= 7*574):(9.43`= )*= 1&= 3:97.9.43= A= 5&79.7= )*= (&7'43*= +43,.6:*= *9= )*= 1&= ,*72.3&9.43_= *8= )*:== 545:1&9.438=8*=).89.3,:*39=5&7=1*:7=548.9.43=,S4,7&5-.6:*=*9=1*:7=(1.2&9_=&=545:1&9.43=)*= 4.,3*;.11*= *89= 8.9:S*= )&38= :3= 84:88'4.8= )*= -U97&.*= &88*?= 842'7*`= 6:.= 54:77&.9= -S'*7,*7= )&:97*8=*85T(*8=2>(4-S9S749745-*8`=9&3).8=6:*=1&=545:1&9.43=)*=439*+*77.*7=(4143.8*=:3*= 5*:51*7&.*=(:19.;S*`=(1&.78*2S*=*9=&88*?=8T(-*`=41=1*8=*85T(*8=2>(4-S9S749745-*8=3*=54:88*39= 5&8_= = 439+*77.*7`= *3= )S':9= )*= 8&.843`= 1*= )S;*1455*2*39= *89= 1*= 2U2*`= *9= 1&= 8*:1*= ).++S7*3(*= &55&7&9= &:= 3.;*&:= )*8= +*:.11*8_= = 5&79.7= ):= 24.8= )*= /:.3`= (*79&.3*8= +*:.11*8= )&1'.348=(422*3(*39=A=8S(-*7`=*397&.3&39=94:9*=1&=9.,*=*9=1*8=+7:.98=6:.=+.3.88*39=5&7=942'*7_= *= 5-S342T3*= 94:(-*= '*&:(4:5= 51:8= 1*8= .3).;.):8= &1'.348= 6:*= 1*8= .3).;.):8= ;*798`= 2&.8= )&:97*8=+&(9*:78=;.*33*39=2*997*=+.3=A=1&=;.*=)*8=+7:.98=)&1'.348`=9*11*=6:*=1&=57S)&9.43_=*8= .3).;.):8= &1'.348= 8439= )43(= 5*:= 342'7*:== A= 54:;4.7= ).85*78*7= 1*:78= ,7&.3*8`= 6:4.6:*= ;.&'1*8_=*5*3)&39`=1&=).85*78.43=):=5411*3=&=1.*:=&;&39=(*8=5-S342T3*8=*9=8*=+&.9=)&.11*:78= *3=2S1&3,*=&;*(=1*8=.3).;.):8=;*798= (+_ &79.(1*= _=&1'.3.82*`=8.1=&=:3*='&8*=,S3S9.6:*=5*:9= )43(=U97*=2&.39*3:*=5&7=1&=).85*78.43=):=5411*3_=*=51:8`=.1=*89=51:8=).++.(.1*=)S1.2.3*7=1*8= ,T3*8=)S1S9T7*8=).85*78S8=5&7=1*=5411*3=45*?= *9=&1_` =,**2=(*=6:.=*=51.6:*7&.9=1*=2&.39.*3=)*= 1&1'.3.82*=84:8=+472*=):3=(7>594541>2475-.82*_= &:97*=5&79`=1&79.(1*=84:1.,3*=1.25&(9=)*8=+&(9*:78=*=9S7.*:78=57S)&9.43`=8S(-*7*88*= 8:7= 1&= 8:7;.*= )*8= .3).;.):8= &1'.348_= = 4.,3*;.11*`= 1*8= .3).;.):8= &1'.348= 3*= 57S8*39*39= 5&8= :3*=9*11*=2479&1.9S=8&.8433.T7*_=*8=(43).9.438=7*88*2'1*39=)&.11*:78=&:=2.1.*:=)*=;.*=)*=1&= 51:5&79= )*8= 2>(4-S9S749745-*8a= :3=*3;.7433*2*39= +7&.8`=-:2.)*=*9= 842'7*_= 1=*89= 5488.'1*= 6:*=1&1'.3.82*=84.9=84:2.8=A=:3*=8S1*(9.43='&1&3(S*`=*9=6:*=843=S;41:9.43=84.9='146:S*=84:8= :3=(1.2&9=9745=8*(`=*=51.6:&39=1*=5*:=)*=2*39.438=)&38=1*8=+47U98=2S).9*77&3S*33*8_= Non-photosynthetic albino variants in Cephalanthera damasonium : a dead-end way in evolution toward mycoheterotrophy?

M. Roy, A. Tognetti, J. Renoult, J.-C. Thomas, L. Misson, B. Schatz, C. Damesin, M.-A. Selosse

Centre d’Ecologie Fonctionnelle et Evolutive, UMR 5175, CNRS 1919 route de Mende, 34093 Montpellier cedex 5, France Laboratoire d'Ecologie, Systématique et Evolution, Université Paris XI, Bât. 362, 91 405 Orsay cedex, France Département de Biologie, Ecole Normale Supérieure, 46 rue d’Ulm, 75230 Paris, France,

This manuscript is in preparation and will be submitted to Ecological monographs .

1 SUMMARY - Albino plants, that do not photosynthetize, usually do not survive into the wild, unless they are supplied with sugars. But some albino individuals were recorded in orchid species, and these albino produce fruits and survive over several years. Such individuals receive their carbon from mycorrhizal fungi and are thus mycoheterotrophs. Green individuals belonging to the same species receive part of their carbon from fungi, and are thus mixotrophs. - These species are closely related to full mycoheterotroph species. Then, such albinos could represent a transition toward complete mycoheterotrophy. Nevertheless, such albinos are always rare and do not invade populations, then may be counter-selected in some ways. Evaluating selection pressure on albinism would help understanding how complete mycoheterotrophy arose, and may underline what adaptations were acquired in mycoheterotroph evolution and that may be linked with photosynthesis loss. - We compared the fitness of albino and green individuals of Cephalanthera damasonium in two different populations, one in oak forest, under oceanic climate, and the second in a cultivated poplar forest in Mediterranean region. The first population looks like mycoheterotroph plants habitat whereas the second never present such mycoheterotroph plants. - Albinos were more sensitive to light stress, pathogen and herbivores. Moreover they have a lower carbon budget than green, and then produce less shoot, less flower, less fruits and less seeds. In all, albino in Mediterranean region had a 64 times lower fitness than green, whereas no difference was detected in Northern region. As only their female fitness is lower, they can still disperse their pollen and reproduce with green. Then, albinism may be maintained as cryptopolymorphism in these species. - Albino are clearly counter-selected and underline the cost of non-photosynthetic leaves, that are reduced in complete mycoheterotroph. Differences between the two sites underline that environmental factors counter select albinism in dry and open area, then mycoheterotrophy may arose only in under storey, and could have evolved from such albino in mixotrophic species.

2 INTRODUCTION

Several eukaryotic lineages became photosynthetic thanks to acquisition of plastids. by endosymbiosis (Archibald, 2009). In further evolution, many lineages became photosynthetic, either with full loss of plastids (such as Oomycetes, Ciliates or Trypanosomids) or with retention of non-photosynthetic plastids (such as Alveolobionts or some Euglenoids and Dinobiontes) (Krause, 2008; Reyes-Prieto et al. 2008; Saldarriaga et al. 2001). The frequent reversion to heterotrophy also arose among Angiosperm, that retains non-photosynthetic plastids. At least 11 families, representing nearly 1% of Angiosperms species (Barkman, 2007), survive either as direct parasites. Other plants receive carbon provided by fungi associating to their roots, forming the mycorrhizal symbiosis. The so-called mycoheterotrophs (MH), are widespread among Angiosperms too, with 400 MH species occurring in 8 families (Leake, 1994). The transition to MH life is frequent among Orchidaceae, where MH nutrition type arose nearly 20 times (Molvray, 2000).

Despite the recurrent loss of photosynthesis in evolution, few models are available on the evolutionary transition between auto- and heterotrophy in plant evolution. Some euglenoids (Euglena spp.) were anciently reported to become achlorophyllous upon various treatments, including prolonged darkness (Rosati et al. 1996), but beside this model, it is often assumed that no other specie present a mix of autotrophic and fully heterotrophic individuals. In photosynthetic angioperms, the apparition of achlorophyllous (= albino) individuals is a frequent event, as shown in mutageneses experiment (that produces 1.2 % of albino mutants in Arabidospis thaliana ; Feldmann 1991) but, except under specific lab conditions ( e.g. Beyrle & Smith 1993), those mutants do not survive in the wild. Albino leaves in forest understorey species are often related to a delayed greening symptom (Kursar 2003) and are not stable over time. Punctual loss of photosynthesis can occur in response to parasitism rusts (Robert et al. 2005) or as an adaptation to drought in Xerophyta scabrida (Tuba et al. 1996), but recover photosynthesis in normal (healthy or wet) condition. In this context, the observation that some orchid species possess two permanent phenotypes, albinos and green (= autotrophic; Julou et al. 2005) furnishes an outstanding model to understand the evolutionary transition from auto- to heterotrophy (Selosse & Roy 2009).

These albinos do survive in the wild, remain albino all along their life and even produce fruits (Salmia 1986, 1989b; Delforge 1998; Abadie et al. 2006; Fig. 1). They all are described from

3 species belonging to the Neottieae orchid tribe species (Orchidaceae), in the genera Epipactis (in E. helleborine, E. microphylla, E. atrorubens, E. purpurata, E. albiensis, E. neerlandica, E. tremolsii, E. distans and E. muelleri ; Salmia, 1986; Delforge, 1998; Scappaticci & Scappaticci, 1998 ; Selosse et al. 2004; M.-A. Selosse pers. obs.) and Cephalanthera (in C. damasonium, C. longifolia, C. rubra, C. kurdica , C. falcata and C. longibracteata ; Mairold & Weber, 1950; Scappaticci & Scappaticci, 1998; Julou et al, 2005; Abadie et al. 2006). These species have close phylogenetic relationship with MH orchids, namely Cephalanthera spp. (C. austinae, C. exigua, C. calcarata, C. pusilla and C. gracilis, Bateman et al. 2005; Pedersen et al. submitted), Aphyllorchis spp. (Roy et al. submitted) or Neottia nidus-avis (Selosse et al. 2002). Yet, they remain rare and seem rather unfit (see below). They are thus ecological equivalents to mutants in genetics, i.e. their dysfunctions may suggest what makes MH strategy successful. They also offer fascinating models for comparing the physiology of autotrophy and MH within similar genetic backgrounds.

These albinos survive heterotropically thanks to their mycorrhizal fungi. Exactly as MH orchids, the green species in which they arise are associated with fungi that are otherwise well-known as ectomycorrhizal fungi (Bidartondo et al. 2004; Julou et al. 2005; Abadie et al. 2006). Moreover, green plants of these species already receive part of their carbon form their fungi too, even if they simultaneously photosynthesize, and are called mixotrophs (MX, Girlanda et al. 2006; Selosse et al. 2006, Selosse & Roy, 2009). This is supported by their unusual 13 C/ 12 C ratio, that is intermediate between full autotrophs and MH plants (Julou et al. 2005; Abadie et al, 2006), supporting that they mix autotrophic and MH nutrition. The contribution of “fungal” carbon, called rate of mixotrophy (Philipps & Greg 2001), can be calculated from their 13 C/ 12 C ratio and potentially ranges from 0 to 100% depending on poorly understood environmental conditions, as well as on their phenotype (green or albino). MX species likely represent a predisposition to the survival of albinos and MH evolution. This raises many questions about apparition and success of albino phenotype in green Neottieae. However, little is known about these albinos because they are always rare in populations ( e.g. , Abadie et al. 2006) and did not raise interest of orchid research beyond report of occurrence. Studies on the three European species where they appear frequently ( C. damasonium , C. longifolia and E. helleborine ) were however performed since the 1980’s. Although the important question of the determinism (environmental or developmental versus genetic) remains unanswered up to now, some major data were obtained. The first result is that the phenotype can change over life span for Epipactis albinos (that may undergo some

4 greening and vice-versa; Salmia, 1989b), but that the phenotype remains stable in Cephalanthera spp. (Delforge 1998; Tranchida et al. submitted), even over 14 years (Abadie et al. 2006). This contrast suggests that, in some Cephalanthera populations at least, the determinism may be genetic, although definitive evidence is still lacking. In these populations, albinos did not aggregate spatially or genetically more than green individuals, suggesting that albinos did not represent a unique lineage and were often closer related to green individuals than to other albinos from the same population ( Tranchida et al. submitted ).

In an investigation of a mixed E. helleborine population in Finland, where albinos were quite numerous, Salmia (1986, 1989b) underlined a reduction of plant height, leaf size and stem thickness, together with a lower flower number that were produced later in albinos, but fruits and seeds seemed similar to green ones. In Cephalanthera spp., albinos and green individuals did not show significant morphological differences (Julou et al. 2005; Abadie et al, 2006). However, in C. damasonium at least, they displayed some physiological differences, such as a higher stomatal conductance, a lower basal metabolism and, obviously, a lower pigment concentration except for carotenoids (Julou et al. 2005). There is agreement that fungal associates of albino and green individuals are the same, both in Epipactis spp. (Salmia 1989a; Selosse et al. 2004) and Cephalanthera spp. (Julou et al., 2005; Abadie et al, 2006), although mycorrhizal fungi tend to be more abundant in albino roots (Salmia 1989a; Selosse et al. 2004; Abadie et al. 2006). In both genera, albinos seen on one year tend to more often not produce shoots on the following year as compared to green individuals (Salmia 1986; Tranchida et al. submitted), due to death or increased frequency of dormancy (a stage were plant survive underground without forming any aboveground shoot, probably thanks to MH nutrition). The most striking feature is that albinos remain rare in populations (Abadie et al, 2006; Tranchida et al. submitted) for unknown reasons. A second potentially important difference is that albinos produce less often fruits (Salmia 1989b), perhaps in part because albino shoots get dry in the summer, earlier than green ones, well before fruit ripening (Julou et al. 2005). This feature, added to the increased frequency of dormancy, may lead to reduced fitness of albinos.

This study is a multidisciplinary comparative analysis of albino and green phenotypes in C. damasonium , a species where the chlorophyll phenotype remains stable (Julou et al. 2005), and for which a population with numerous albinos was recently discovered in South of France (Tranchida et al. submitted), thus allowing destructive analyses and replicates of independent

5 measurements. Whenever possible, comparative studies were made in other populations containing albino. To verify and try to elucidate the origin of differences between phenotypes, we carried out a large array of investigations on both phenotypes, covering the whole lifespan of the shoots and thus, focusing on traits related to vegetative life (shoot development and physiological features) and reproduction (flowering, mating, fruiting, seed germination). We especially focused on the origin of C in green individuals (fungal vs. photosynthetic) at various times in the season and for different organs; on the susceptibility to biotic stress (pathogens and herbivores); on the difference of fitness for albino and green shoots; and on factors explaining the drying phenomenon, such a stomatal exchanges. We aimed at a global picture of albinos’ morphology and physiology, compared to green individuals, to find out different (possibly linked) reasons explaining the differences observed, and the reasons for limited albino success in natura . Considering albinos as an evolutionary step, perhaps failed, in MH evolution, we discuss from these results what makes an emerging MH plant successful. We thus try, in the discussion, to unravel important adaptations in the evolutionary pathways to MH plants.

MATERIAL AND METHODS

Sampling sites The whole study was handled in Montferrier (MON; Tranchida et al. submitted), and part of the study was pseudo-replicated on another C. damasonium population at Boigneville (BOI; Julou et al. 2005). Isotope measurements were also done on green individuals of three other species: C. longifolia and Epipactis helleborine in Fontaret (43°54’52’’N, 03°30’23’’E, elevation 400 m) and C. damasonium , C. rubra and E. helleborine at Chauriat (45°45'25”N, 03°17'21”E, elevation 500 m); only green individuals occurred at these sites. MON is a poplar plantation nearly 15 years old that was a wineyard until the 1990’s. Populus nigra are planted every seven meters, and young Quercus pubescens are growing in between. The soil is argillaceous, covered by Clematis vitalba and Hedera helix , and the field is flooded sometimes by the Lez river in spring and fall. BOI is described in Julou et al. (2005).

Morphological traits The MON population was discovered by us in late spring 2006 and investigated in 2007 and 2008. Only some of the individuals were spotted in 2006, but all of them were recorded in

6 2007 and 2008, to investigate their position (Tranchida et al. submitted) and morphological parameters. In 2007, shoot apparition, growth, flowers and fruit development, herbivory signs and invertebrates occurrence were recorded every week from March to end of June. In 2006 and 2008 in MON and in 2007 and 2008 in BOI, morphological parameters of were recorded in June only. We focused our study on aboveground shoot and reproduction apparatus, since Julou et al. (2005) already investigated mycorrhizae at BOI. For morphological study, parameters were measured in the field using a Vernier Caliper. Morphological parameters were focused on the shoot (shoot number, height, reproductive height, dried shoot or absence), on the leaves (leaf number, first leaf length and width, herbivory signs, infection spots, presence of insects, drying signs or absence), on the flower (flower number, number of flowers present, fallen, eaten, presence of insects and pollinia removal) and on the fruit (fruit number, aborted or fallen fruit number, ovary length and width). To avoid destroying the population, only leaves collected for isotope sampling were used to measure the surfacic leaf area (SLA) by weighting and measuring their surface when they were fresh and weighting them after 3 days drying at 65°C (in MON only). Leaves were collected in 2006 to investigate stomata density on fresh leaves (n=10 per phenotype, in MON only) by separating cuticle from the leaf with thin tweezers and then observing it under a 40X microscope. For each leaves, 10 countings of stomata number at X400 were performed, randomly located on the leaf. Nearly 1 cm² of each of these same leaves (collected at the edge) were fixed in gultaraldehyde for transmission electronic microscopy as in Roy et al. (2009). Leaves fixed were embedded and manual transverse sections (10 µm) were cut with a microtome, differentially stained with a mixture of safranin and fast green FCF (Bryan, 1955), rinsed with distilled water and observed with a 40 X light microscope to measure leaf and cuticle thickness. Susceptibility to herbivores and pathogens Counting of insects and pathogens infection (black spots on leaves) resulted from observations but were not in the experimental scheme at the beginning. Herbivores were identified when feeding on the plants, such as criquets, caterpillar, ants, and most of them were installed permanently on leaves or shoots. Herbivory signs were observed on leaves, flower and fruit and consisted of absent part of the organ delimited by an irregular border and not attributed to necrosis. Other invertebrates as mites or arachnids were recorded on flowers or fruits only and aphids were recorded on fruits only, feeding on it. Pathogens infections were identified as black spots on leaves. A few leaves were observed under microscope to observe these spots.

7 Pigments measurements and analysis Leaves of green (n=5 per site) and albino (n=5 per site) individuals were collected in 2006 from MON and BOI, measured, weighed and kept with carbonic ice until reaching the lab, where they were kept frozen at -80°C until use. Entire leaves were grinded in liquid nitrogen and pigments were extracted in 80% acetone:water. Absorbance spectra were obtained with an Aminco Spectrophotometer to measure chlorophyll a and b quantities, calculated as in Julou et al. 2005. For two leaves of each phenotype, detailed carotenoid composition was investigated by High-Performance Liquid Chromatography.

Gas and water exchange physiology Gas and water exchanges were measured punctually at various Photosynthetic Active Radiation (PAR) levels, i.e. at PAR=0 or PAR=100 µmol m -² s -1 , representing the light condition in the site, when leaves were mature i.e. when individuals were blooming in May or later in June, in 2006, 2007 and 2008 in MON. They were also measured in response to variations of PAR (Photosynthetic active radiation), CO 2 and Vpd (Vapor pressure deficit) variations, both in May and June in 2007 and 2008 in MON and in June in 2006 and 2007 in BOI. We used an infrared gas analyser (LI 6400, Li-Cor, Inc., Lincoln, NB, USA) at constant

CO 2=300ppm; measures were done between 8:00 and 11:00 a.m. to avoid high temperature and stomata closure. The Li-Cor used in 2007 at MON only produces red-light whereas the Li-Cor used in BOI, and in 2008 in MON includes both red and blue light. Entering air was pumped two meters away from the plant studied, in a 50L recipient, to avoid interferences with vegetation around. Each measure is a mean of ten records in a single minute. Measures were spaced by at least 5 minutes to allow leaf adaptation to the new PAR conditions and to check H 2O and CO 2 reference stability; measures were recorded only after 20 minutes of leaf adaptation to the chamber. Leaf temperatures were measured punctually with an Optris Minisight Plus infrared thermometer in MON only in 2006.

Isotopic sampling and analysis Experiment 1. Seasonal variations in carbon nutrition Seasonal variation of 13 C and 15 N of green C. damasonium leaves were investigated in MON by collecting leaves at the end of March (shoot apparition), the end of April (full development, bud), mid-May (blooming time), the end of May (fruit formation), mid-June (one month old fruits) and the end of June (beginning of fruit dispersal). For each sampling,

8 we harvested n=5 leaves for green C. damasonium , Clematis vitalba and Rubia peregrina . The sampling site was the same all along the season (a 7x7 m square). To avoid destroying too many albino, n=5 leaves were collected only in late June to get a mycoheterotrophic reference. As no fungal fruitbody was seen during the spring, Inocybe cf. dulcamara (n=5) were collected in November 2007 on the same site. To check for fungal colonization at fruiting time, when albino are drying, n=3 root systems for both albino and green individuals were harvested. Colonization rate was measured by differentiating roots infected and not infected by their colour, followed by thin cuttings and pelotons observations every centimetre. The surface was calculated using length and diameter of root colonized.

Experiment 2. Comparison of heterotrophy level between organs Inter-organs variations of 13 C and 15 N in green C. damasonium were investigated in three different sites (MON, Fontaret and Chauriat) but not in BOI due to the low number of plants (Julou et al. 2005 already reported 13 C and 15 N for the leaves) and included other Neottieae species for comparison. In MON, n=5 fruits (albino and green) were collected in late June together with n=5 green and albino leaves (from the first experiment). In Fontaret and Chauriat, n=5 C. damasonium , C. longifolia , C. rubra or E. helleborine were collected and organs were separated (roots, rhizome, shoot, leaves, flower or fruit). For controls, n=5 leaves of 2 autotrophic reference species, aerial parts of mycoheterotrophic species were harvested. For both experiments, All green leaves were collected all at same light level and same distance form the ground (less than 0.5 m) as orchids, to avoid 13 C distortion due to higher photosynthetic rate or CO 2 resulting from soil respiration. Samples were dried at 65°C for 72 hours and handled as in Tedersoo et al. (2007) to measure total N, C/N and abundances of 13 C and 15 N. Isotope abundances are expressed in 13 C and 15 N values in parts per thousand 13 15 relative to international standards V-PDB and atmospheric N 2: C or N = (R sample / 13 12 15 14 Rstandard – 1) x 1000 [‰], where R is the molar ratio, i.e. C/ C or N/ N. The standard deviation of the replicated standard samples (n=13) was 0.037 ‰ for 13 C and 0.197 ‰ for 15 N. Mixotrophy rate was calculated using a two-source mixing model (Phillips & Gregg, 2001), with mean 13 C or 15 N values of autotrophs and MH or albino plants as references for, respectively, autotrophic biomass and mycoheterotrophic (fungus-drived) biomass.

Odour sampling and analysis Odours were collected in May 2007 at MON on n=5 albino individuals with 2 to 4 opened flowers each and n=8 green individuals with 2 to 5 opened flowers each, using adsorption-

9 desorption method (dynamic head-space technique) as is Proffit et al. (2008). Briefly, upper shoot and opened flower were enclosed in polyethylene terephtalate (Nalophan®) bags paper, closed with wool strings and cotton to avoid air leaks. The pumped air was collected with trap containing 30 mg of Alltech Super Q absorbent (ARS Inc., Gainesville, FL). Odour sampling was performed during three hours, between noon and 3:00 p.m. Ambient air was also collected in an empty bag as a control (one control per site and sampling date). Trap were kept on dry ice until reaching the lab and latter at -20°C and were later handled as in Proffit et al. (2008) to elute volatiles. Volatile compounds were analysed by injection in a CP-3800 (Varian Inc., Palo Alto, USA) gas chromatograph with FID detector coupled with a Saturn 2000 mass spectrometer (Varian Inc., Palo Alto, USA), together with two internal standards (nonane and dodecane, 200 ng )l–1). Pollutants were removed from the analysis by comparing volatiles from orchid and those from empty bags. Volatiles quantities were calculated by reference to the internal standards as in Proffit et al. (2008) and divided by the number of opened flowers.

Pollination experiments On May 2008 in MON, when flower began to open, pollinia were cut and used to pollinate either a flower from the same inflorescence (geitonogamy), or a flower from a different individual – in the later case, the second individual had either the same phenotype (homogamy) or a different phenotype (heterogamy). Then inflorescences were protected from other pollinations by mesh bags closed with cotton and wool strings, and tutored by a wood stick to avoid shoot injuring. Inflorescences already pollinated and unpollinated flower buds were also enclosed to obtain respectively controls without herbivory or mesh bag bias of natural pollination (random mating) and of natural autogamy. In July 2008, fruits were collected, measured, weighed and then dried 72h at 65°C and re-weighted. Some fruits were already opened and probably had lost some seeds at collecting time. Thus, fruit weight are unreliable measures of seed number, whereas measures based on fruit size are better estimators of seed set (before opening, fruit volume correlates with seeds number and total seeds weight – data not shown). As fruits were dried when collecting, they kept their size and fruit length and width were thus measured and used as one indirect fitness estimator.

Germination experiment On late-July 2006 in MON, fruits were collected. Ten albino fruits and 10 green were sampled for seed counting and observation. Forty one albino fruits and 50 green were collected and

10 two seeds pool (from albino and from green individuals) were prepared for sowing. Circa 200 seeds were enclosed between two 4x4 cm of nylon cloth (50µm pore size) fixed by two parts of camera slide. The resulting 91 mesh bags were dried with silicagel during one month, to mimic natural conditions of fruit drying, and buried in the vicinity of mother plants at 5-7 cm depth as in Van den Kinderen (1995). They were randomly spread on a 1x1 m grid, spaced by 10 cm from each other. Seed packet were recovered 18 months later, in January 2008, washed with sterile water and observed under a dissecting microscope (60X) on the same day. Ungerminated seeds were assigned to four stages: (A), no proembryo formed with decayed testa; (B), intact testa containing a proembryo; (C) swollen pro-embryo with cracked testa; protocorm with trichomes; (D). As 18 month seemed to early for germination, successful germination was here defined by stage C and D and may reflect only the first steps. Some ectomycorhizae had grown inside the seed packet, and both protocorms and ectomycorrhizae were isolated for molecular investigations and kept at -20°C.

Molecular investigations Fungi of leaves presenting spots and fungi colonizing protocorms were identified by molecular investigation, together with ectomycorrhizae that had grown inside the buried mesh bags. Extraction, PCR amplification of the internal transcribed spacer (ITS) of the fungal ribosomal DNA, and sequencing were performed as in Roy et al. (2009). Then, edited sequences were deposited in GenBank, and compared with GenBank databases (National Center for Biotechnology Information, http://www.ncbi.nlm.nih.gov ) using Blast (Altschul et al. 1997) to infer the fungal identity.

Statistical analysis on isotopic measurements Morphological dataset and germination dataset were analyzed either by comparing each trait between phenotypes by Student test of by performing a component factorial analysis, using Minitab TM . Other parameters, obtained from few individuals only (total N, C/N, 13 C and 15 N, pigments concentrations), were compared between phenotypes by Mann-Whitney tests using Minitab TM . Total N, C/N, 13 C and 15 N studied along the season were also compared by an ANOVA and then independently between species on each site by pairwise Mann- Whitney tests. Physiological data were compared either for each measuring point between phenotype by Student t-test or for each phenotype between points by an ANOVA. Results of artificial crossings were compared between phenotypes and treatments by an ANOVA.

11 RESULTS Population dynamic between years At MON, the number of shoots recorded in 2007 and 2008 was higher than in 2006. The percent of albino shoots among the population decreased over the three years, from 23.8% to 6.4%. Most of individuals of both phenotypes produced shoots only on one year and a few produced shoots over three years (Table 1). The mean life span per individual was not significantly higher for green (according to an ANOVA, F=0.05, P=0.815). More dormant individuals (shoots seen in 2006 and 2008, individual persisting hypogeously in 2007) were recorded for albinos than for green individuals, resulting in nearly 30 more dormant albinos than green in proportion (Table 1). At BOI, the number of shoots recorded in 2006, 2007 and 2006 was nearly constant, with a ratio of albino to green individuals higher than in MON. As in MON, over the three years, albino proportion decreased from 62.5 to 47.3%. Although only albinos produced shoots over three years, more albinos were dormant or occurring on a single year, as in MON. The mean life span was not significantly different between phenotypes (ANOVA, F=1.1, P=0.3). In all, albinos tend to produce less often shoots and tended to decrease in frequency over the three years.

Shoot development over the year At, MON, first albino individuals were recorded in 2007 from March 26th, like green ones, and new shoots appeared until June for both. Time of shoot appearance did not differ between phenotypes (ANOVA, F=2.10, P=0.242, Table 2, Fig. 2). Flowers began to open in May too, significantly earlier for albinos (ANOVA, F=8.49, P=0.04, Table 2) that bloomed younger than green individuals (ANOVA, F =6.36, P=0.014, Table 1). Similarly, albinos fruited earlier (ANOVA, F=12.49, P<0.0001) and younger (ANOVA, F=13.44, P<0.0001, Table 2), suggesting a shorter duration of reproduction. A few individuals, mostly albino, presented necrotic drying of leaf, starting from the edge, starting from May but mostly in June, and albinos always dried before green individuals (Fig. 1 ; Table 2 ; ANOVA, F=11.8, P=0.001). These symptoms amplified in June and for part of affected individuals, the whole shoot dried and disappeared a few days after. This drying phenomenon did not occur only albinos, but was recorded more often and earlier in shoot life for albinos (Table 2), so that half of the albinos were lost before fruit opening and seed dispersal ( versus 7.6% of green individuals, Table 2). Similar observation were made at BOI in early June 2007, where 3 of 13 albinos dried but none of green individuals (no drying happened in 2008; data not shown).

12 Vegetative morphology No significant difference in general morphology was seen at MON, but albino shoots tended to be smaller (only significant in 2007, ANOVA, F=15.23, P<0.001, Table 3). Albino leaves were similar in numbers as compared to green ones, but appeared much less developed (ANOVA for leaf surface: F=5.41, P=0.026; for leaf thickness F=338.91, P<0.0001) with thinner cuticle (ANOVA, F=55.35, P<0.0001), a lower stomata density (ANOVA, F=61.6, P<0.0001, Table 3); they contained less chloroplast than green ones (Fig. 3). Chlorophylls a and b were 100 times less concentrated in albino leaves (Mann-Whithney test, P=0.0012), and the total quantity of pigment was 60 times lower (Mann-Whitney test, P=0.02, Table 3), even when chlorophylls were not taken into account (Mann-Whitney test, P=0.02). Carotenoids were 4 times less concentrated in albinos (Mann-Whitney test, P=0.02, Table 3), and pigments composition did not reveal light-stress carotenoids, such as zeaxanthine (Table 3). Similar trends were observed at BOI in 2007: albino shoots were three times smaller than green ones in 2007 (ANOVA, F=10.54, P=0.005), and not significantly higher in 2008 (ANOVA, F=0.7, P=0.419); both phenotypes produced a similar number of leaves (Table 3). As in MON, albino leaves contained 100 times less chlorophylls, 4 times less carotenoids, and no light-stress carotenoids (Table 3).

Comparison of gas exchanges Leaf physiology was investigated, especially in order to understand the C metabolism and drying phenomenon. Basal metabolism at MON, as established by dark respiration, was 2 to 4 times higher in green leaves on every experiment (For each experiment, P<0.0001, Fig. 4); no photosynthesis was detected in albino leaves, although the net assimilation was significantly different, but not always higher, when PAR increased (ANOVA, F=19.13, P<0.0001). Net assimilation was always significantly higher for green individuals (ANOVA, F=8.52, P<0.0001, Fig. 5) and positively responded to PAR increase (Fig. 5). Compensation point for green individuals changed from PAR=20µmol.m -2 .s -1 during blooming (May 2007) to 9±0.5µmol.m -2 .s -1 during fruiting (June 2007 and 2008; not shown). The same trends were observed in BOI. Basal metabolism was 3 times higher for green leaves that positively responded to PAR, with a compensation point at PAR=10µmol.m -2 .s -1 (Fig. 6).

In MON, in all light conditions, stomata conductance was 2.5 to 5 times lower in albinos than in green individuals (ANOVA, F=1073.03, P<0.0001, Fig. 4). Using blue and red light, reaction to PAR increase showed significant differences of stomata conductance, but no

13 correlation with PAR values in albinos (Fig. 5), whereas green ones presented a typical increasing stomata conductance to PAR increase for PAR>200µmol.m -2 .s -1 (for PAR<200µmol.m -2 .s -1 , stomata responded in the opposite way, perhaps due to transient oscillatory periods longer than expected). Measurements performed on blooming individuals (May 2007) showed a significant effect of phenology within each phenotype, with a higher stomatal conductance in May, but again 5 times higher for greens (ANOVA, P<0.001 for each phenotype, Fig. 4). Using red light only, no response to PAR variation was observed (Fig. 7). Albino leaf temperature was not significantly higher than for green ones, when measured punctually at ambient light in 2006 (27.970±0.15 (n=5) for albino vs. 26.87±0.98 (n=10) for green), but when increasing PAR, green leaf temperature at the beginning of the experiment were cooler than for albinos (ANOVA F=124.96, P<0.0001, Fig. 5) and increased quicker than albino one with PAR (R²=15.7%, P<0.0001 for albinos vs . R²=93.9%, P<0.0001 for green individuals, Fig. 5). Albino in BOI showed a lower conductance in 2007, but only 20% lower than green and higher than in BOI (Fig. 4) and even negative and 5 times lower in 2006 (ANOVA F=89.16, P<0.0001, Fig. 4). Albino leaf temperature was significantly higher than green one but showed a lower warming rate than green (R²=23.3% P<0.0001 for albino vs. R²=89.9% P<0.0001 for green, Fig. 4), as in MON. Outside air temperature was around 17°C in BOI and humidity rate around 60%, then conditions are difficultly comparable with MON data, were temperature was ranging from 25°C to 30°C with 30% of humidity. For both phenotypes, stomata responded to humidity rate changes, and conductance was lower for very low humidity rate (<15%) and high ones (>55%). In all, albino had always a lower basal metabolism and a lower conductance. Temperature seems to depend more on the outside conditions and light environment.

Seasonal variations in carbon and nitrogen nutrition In order to estimate the contribution of fungal carbon to above ground biomass over a growing season, samplings were performed in 2007 over the growing season at MON. 13 C values of green C. damasonium were always significantly higher than that of autotrophs (Mann-Whitney tests for each date always significant, P<0.05), but not always significantly different from 13 C values of Inocybe cf. dulcamara and of albinos (Fig. 8a), suggesting changes in mixotrophy rate among the vegetative season. 15 N values of green individuals were always significantly higher than for autotrophs and lower than for albinos (Mann- Whitney tests per date always significant, P<0.05; Fig. 8b), as expected for mixotrophic individuals. 13 C and 15 N values of reference species for autotrophy ( Rubia peregrina and

14 Clematis vitalba ) did not significantly different each date (Mann-Whitney tests per date, P>0.05) nor between sampling dates (ANOVA, F=2.40, P=0.099 for C; ANOVA, F=1.33, P=0.326 for N; Fig. 8). Albinos, that had significantly higher 13 C values than Inocybe cf. dulcamara (Mann-Whitney test, P=0.034), but did not significantly differ in 15 N (Mann- Whitney test, P=0.557) as expected for mycoheterotrophic species. They were thus used as a mycoheterotrophic reference: since C. damasonium 13 C decreased continuously from -24‰ to -28‰, heterotrophy rates showed significant differences between dates (ANOVA, F=21.97, P<0.0001), and continuously decreased going from 80% in March to 20% in June (Fig. 9). C/N of green C. damasonium leaves was significantly higher than autotrophs and lower than albino mycoheterotrophic reference (leaves or fruits; Mann-Whitney tests always significant, P<0.05; Fig. 9). Albino leaves C/N was four times lower than for green leaves (Mann- Whitney P=0.0018), while albino fruits were 40% lower than green fruits in C/N (Mann- Whitney P=0.02). C/N values of green C. damasonium leaves increased significantly during the season (ANOVA F=2.22, P=0.097, Fig. 9). Since C/N is supposed to decrease with increasing heterotrophy level (Julou et al. 2005), this confirms a change in nutrition over the season, toward more heterotrophy, in congruence with a higher net assimilation rate at fruiting time (Fig. 4). Roots were investigated only once in June 2007 (fruiting period) but the colonization rate was not different between green and albino individuals (nearly 60% of root surface was colonized, Mann-Whitney test P=0.67), showing that plants were still mycorrhizal in June and albino individuals had the same quantity of fungi.

Comparison of heterotrophy level between aerial organs We tested the possibility of different levels of heterotrophy between fruits and leaves, using additional samplings at Chauriat and Fontaret. At Montferrier, 13 C of C. damasonium green fruit and leaves were not significantly different (Fig. 10). At Chauriat and Fontaret, the orchids revealed significant 13 C and 15 N differences between autotrophic and heterotrophic organs (Fig. 11, 12, pairwise Mann-Whitney tests always significant, P<0.05). Differences of 13 C and 15 N between roots and shoots were about 4‰. Assuming that 13 C variations only depend on the level of use of fungal C, using green leaves of Hedera helix and Trifolium arvense as a reference for autotrophy and mycoheterotrophic plants ( Neottia nidus-avis ) as a reference for heterotrophy, the calculated percentage of fungal carbon in leaves varied between 39 to 47% at Fontaret, and differences between organs were often significantly different in the order root > rhizome > shoot > fruit > leaf (Fig. 13). At Chauriat, evidence for

15 similar trends in three related orchid species ( Cephalanthera damasonium , C. rubra and Epipactis helleborine ) were observed, leaf percentage fungal carbon varied between 45 and 83% and differences between organs were significantly different and ordered as in Fontaret (Fig. 13). Then, fruits appeared to be significantly more autotrophic than shoots or non-green parts, so that in green individuals, photosynthetic C is important for fruit biomass (20 to 63% of fungal carbon vs. 43 to 100% of fungal carbon, Fig. 13).

Susceptibility to herbivores and pathogens At MON in 2007, signs herbivory were observed mostly on leaves, rather than on flower or on fruits (Table 4) for both phenotypes, mostly on albino leaves (43% vs. 15%, ANOVA F=21.86, P<0.0001). At BOI, more signs of herbivory were observed in 2008 (ANOVA F=8.78, P<0.012), although an opposite, non-significant trend occurred in 2007 (ANOVA F=2.67, P<0.178; the low number of individuals limits the analysis here). At MON, all recorded herbivors were significantly more frequent on albino (P<0.01 at least, Table 4), and 31.6% of albino bared at least one herbivorous invertebrate vs . 3.7% of green individuals (ANOVA F=12.85, P<0.0001). As far as reproductive organs are concerned, albino fruits attracted more aphids and ants (ANOVA F=4.06, P=0.044) and albino flowers were more frequently eaten (ANOVA F=9.63, P=0.002), although the aphid load on fruits did not differ (ANOVA F=1.23, P=0.268).

In June 2007 and 2008, at both MON and BOI, leaves of both phenotypes presented black spots (Fig. 14), accompanied or not by leaf drying, that were identified as Pseudocercopsora sp. by ITS sequencing (to be deposited in Genbank). Since non-drying leaves sometimes had infections, and some drying leaves were uninfected (not shown), the drying seemed independent of to the fungal infection. Infections symptoms were significantly more frequent on albinos at MON in 2007 (55 vs. 35%; ANOVA, F=371.98, P<0.0001; no data in 2008). At BOI, both phenotypes were equally infected by Pseudocercopsora sp. in 2007 (ANOVA, F=0.19, P=0.668), whereas in 2008, only albinos were infected (Table 4). Nevertheless, infection rates at BOI were reduced as compared to MON. In all, albino appeared more sensitive to leaf herbivores or leaf parasites than green individuals.

Artificial crossing

16 In experimental crosses, no significant effect of type of crossing at MON was revealed for green individuals or albinos, but albino fruiting success was significantly 4 times lower than green ones in all crosses (ANOVA F=44.0, P<0.0001, Fig. 15). No cost of forced geitonogamy was detected in both individuals; interestingly, enclosed buds did not produce any fruits, suggesting that spontaneous autogamy do not occur, at least under these conditions. The fact that fruit size does not significantly on pollen source demonstrates an important effect of mother phenotype on seeds number.

Natural pollination Pollination as seen by pollinia removal occurred for both phenotypes, but higher for green individuals (Table 5, ANOVA F=11.81, P<0.001). However since pollinia are powdery (Pridgeon et al. 2008), pollination can occur without pollinia removal. Unfortunately, no pollinator was identified, despite hours of in situ sampling and study. Nevertheless, the occurrence of predator spiders ( Tomis ) near flowers suggests that some insects visit flowers. Pollinia of flowers opened since 3 days often pended without falling (Fig. 16), then autogamy could occur before ripening if any insect visit the flower. The volatiles quantity was very low for both phenotypes and neither significantly different from the controls (ANOVA, F=2.01, P=0.165, not show) nor different between each other (ANOVA, F=0.18, P=0.681, Table 5). Nevertheless, the bouquet seemed rather different in quality, and among 7 component specific to C. damasonium , none was detected only in albino flowers whereas four where restricted to green flowers (to be identified). The occurrence of green specific volatiles could explain the higher green flower pollination, but components have to be identified to check for their role in putative pollinator attraction.

Fruiting and seed set For both phenotypes at MON, flower number per shoot significantly correlated with shoot size (Table 5). In 2007, albino produced less flowers (ANOVA F=8.49, P=0.004, Table 5) and less fruits (ANOVA F=12.49, P<0.0001) but with similar ratios of fruits per flower (F=0.86 P=0.354), so that differences in fruit set resulted from differences in number of flowers. Fruit were two times smaller in length and width for albinos (Table 5, ANOVA, P<0.001). Albino fruits contained 4 times less seeds (ANOVA, F=58.22, P<0.0001). Seed viability was not tested, but the percentage of seeds with embryo was similar (ANOVA, F=0.54, P=0.465). After 18 months underground, albino and green seeds produced very few protocorms, in identical proportions (nearly 0.7% of seeds per packet, ANOVA, F=0.12, P=0.727, Table 5),

17 but 4 times more green seeds presented first signs of germination (stage C, ANOVA, F=94.47, P<0.0001). Thus, albino seeds seem able to germinate, but are 4 times less numerous and 4 times less viable than in green fruits; and fruit are less numerous (but not always, it depends on the shoot size) and half of the shoot dried before seed dispersal: in all, fitness seems at least decreased by 64 times.

In 2007 at BOI, albino also produced fewer flowers and fruits (Table 5), with a high (90%) ratio of fruits per flower that was similar in both phenotypes (ANOVA, F=0.09, P=0.778), and fruit size did not significantly differ (P>0.05 for both fruit length and width). In 2008, neither the flower number, nor the fruit number, nor the ratio of fruits per flower significantly differed between phenotypes (P always>0.05 for each test). The BOI population thus show a less pronounced trend to reduced fitness in albinos (seeds viability and germination was not estimated in this case).

DISCUSSION A lot of differences between albino and green phenotype were recorded in this study (Fig. 17), both for vegetative development and reproductive potential. As far as the relationships between albinos and green individuals is concerned, they do not show any sign of reproductive isolation: population are mixed spatially and do not show any aggregation (Tranchida et al. submitted); their phenology is rather similar as their blooming periods are overlapping (Fig. 2), which allows crossing between phenotypes; and controlled inter- phenotype crosses produce similar fruit set to homogamous ones (Fig. 15; the seeds fitness remains however not tested so far). Even if C. damasonium is claimed to be mainly autogamous (Scacchi, 1990), we found little evidence for autogamy without manual pollination in this study (Fig. 15), and its genetic diversity does not strongly differ from the autogamous C. longifolia (Tranchida et al. submitted). Thus, no isolation between phenotypes is detected, confirmed by an overall rather similar morphology and phenology.

Albinos fitness In BOI, female fitness did not differ and they represented half of the population. In MON, where albino represented ca 15% of the population, albino fitness was lowered by shoot

18 drying and herbivory focused on albinos (Table 4 and 5, Fig. 1), as only half of the shoot survived until fruit were mature. Nevertheless, in both population, albinos produced flower and fruits and seemed able to reproduce, as seen in germination experiments at MON. Although the mean life span over 3 years was not significantly different between phenotypes and populations, a higher frequency of dormancy in albinos (30 times more than green individuals) result in a at least a 64 times lower female fitness for albino in MON; taking into account that half of albino shoot dry before seed dispersal, that they produce 4 less seed number (due to 2 times smaller fruits) and the 4 times less seed germination, 64 time lower female fitness is resulting (Fig. 17 and Table 7). Thus, individuals that behave as albinos reproduce less – if any genetic determinism to albinism exist, it is likely not selected. Even if albinos can be numerous (up to 111 over the three years in MON) and reappear over years, their higher dormancy of course explains why they accumulate slower than green individuals in populations. Dormancy is common among terrestrial orchids, and considered as a repair stage where the plants produce no shoot (Rasmussen, 1995, 2002). For example, Listera ovata , another Neottieae, is often dormant the year after producing flowers (Brzosko, 2002) but reappears a few years later. For non-mixotrophic species, at the opposite, dormancy seems more costly, as they may not receive enough carbon from fungi during their underground phase (Shefferson et al. 2003). Moreover, there is evidence that dormancy is a response to stress, e.g. in C. longifolia (Shefferson et al. 2005), so that the increased dormancy may somehow reflect physiological problems of albinos (see below). Since there is no phenotype change (Abadie et al., 2006; Tranchida et al. submitted), albino remain non- green and thus have a lower probability of appearing above ground on a given year as compared to green ones. Thus, even by assuming that albinos and green individuals have similar rate of settlement, the albino shoots number will grow slower than the green shoots number – additionally, female fitness investigation suggest that albinos settlement could be lower – however, we fully ignore the genetic determinism and thus, whether albino offspring are, all or in part albinos. All the factors counter-selecting lowering the fruit production (Fig. 17) occur in June, when fruits are maturing whereas in May, albino shoots are still very similar to green ones in MON, and produce an equal number of flowers. If they produce a same quantity of pollen, and if pollinia / pollen removal is similar in both phenotypes, then albinos may have a male fitness similar to green individuals. It is unclear from our data whether pollinia / pollen removal is similar or not but, given that the ratio of fruit per flowers is identical and that so spontaneous autogamy was not detected, we can assume a similar frequency of insect visit in all

19 phenotypes. Thus, similar level of pollen transmission may occur – but this question remains unanswered. In case of genetic determinism, albinism could transmit through pollen: although maternal inheritance of plastids can be expected (Birky 1995), more than 95% of genes encoding plastid proteins are nuclear-encoded (Krause, 2008; Archibald, 2009), so that albinism, if genetically determined, has a high probability of being be nuclear-encoded. Indeed, our controlled cross suggest that crossing between albino and green are allowed. If albinims is a loss of function, thus recessive, this opens the possibility that numerous individuals are heterozygous, and perhaps not counter-selected: albinism, although unfit, would in this case be difficult to erase (Lopez et al. 2008). However, we ignore whether albinism is a genetic trait maintained by repetitive mutation, or transmission through heterozygous individual (cryptic polymorphism), or even simply a stable developmental state that can affect any individual, upon unclear determinism.

The carbon budget in albinos : less carbon of course… There is evidence that albinos are carbon-limited, since their basal metabolism was lower (Fig. 4). This was already observed by Julou et al. (2005) at BOI, and is probably due to their lower C availability (Williams & Farrar, 1990; Amthor, 1995), as they do not photosynthesize. The lower pigment content and lack of photosynthesis (Fig. 1) may not be fully compensated by the fungal C. Indeed, although previous studies on other populations and species suggested a higher fungal colonization (Salmia 1989a; Selosse et al. 2004; Abadie et al. 2006), a similar colonization level in June was observed at MON, which was rather low (about 60%). This low colonization suggests that albino are lacking carbon in June, while photosynthetic carbon complete the green carbon budget. Comparisons between organs reveal differences in 13 C that are two times higher than inter-organ differences in other plants (Badeck et al. 2005), thus suggesting that fungal carbon may in part drive these differences. Similarly, green leaves show variations in 13 C (-4‰ in 4 months, ending to a value identical to that of fruits), not observed in control plants (Fig. 8), that likely reflect variation of respective contribution of photosynthetic and fungal C. Such decrease of 13 C is described for trees (Damesin & Lelarge, 2003; Maunoury et al. 2008) and is attributed to phenology: plants use more starch during growth period (Damesin et al., 1998). For trees, variations can reach - 4‰ but they are measured on trunks, whose 13 C varies more than leaves one, as it is partly

20 heterotroph and has both starch signature and photosynthesis signature. As leaves have only photosynthetic carbon signature, their variations are lower than trunk or shoot one. As we studied variations on leaves, such decrease is even more significant. Seasonnal variations on herbaceous species (Smedley 1991) are lower (about 1 to 2‰) and the 13 C of our autotroph control only decreased by 2‰. Then, part of the 13 C variation in green C. damasonium leaves can be attributed to phenology but part of this variation may be attributed to changes in fungal carbon flow. Assuming that 13 C variations only result from this, there is a continuous decrease of use of fungal C, from 80% in March at shoots emergence to 20 % at fruits ripening in MON. In fact, the increase in C/N (Fig. 9) is congruent with a trend to more autotrophy. Green flowers and fruits are thus composed of both photosynthetic and fungal C (although fungi are still present in roots in June; Table 3). The fact that albinos have no photosynthetic C may explain some under-performing at late stage of shoot development, such as a lower number of flowers, smaller fruits (Table 5). It may also explain (see above) the increased frequency of dormancy.

Overheating, herbivory, pathogens and fruitset: the tale of a dawn Shoots morphological parameters were similar for both phenotypes (as in Julou et al. 2005), but albino had thinner leaves, with reduced cuticle and stomata density, as shown for E. helleborine albinos (Salmia, 1989a). Their stomata conductance is 2 to 5 times lower than green one. This is opposed to our previous measure at BOI (Julou et al. (2005) that we were not able to reproduce here over 2 years. The lower conductance can be attributed to the lower stomata density and mainly to the lower leaf water content, as leaves are thinner; the thinness of cuticule is not leading to higher conductivity but cuticle chemistry rather than cuticle thickness is important (Kerstiens 1996): the thickness may rather be a barrier against herbivores and invaders (see below). Similarly, change in stomata density does not necessarily in itself the conductance (Berger & Altmann 2000). Investigations of the shoot vessels and water flow should be done to understand better how this lower conductance affects the water budget in albinos.

Albino stomata were functional, as they responded to PAR, CO 2 and vapour pressure deficit variations, as did the green, but always with lower values of water and gas exchanges. It is likely, however, that the lack of chloroplast in stomata may in part modify responses to light. This especially applies to red light response (stomata opening at high light intensity) that depends on photosynthesis and green plastids (Hashimoto et al. 2006; Shimazaki et al. 2007).

21 Albino Vicia faba individuals and albino patches in Chlorophytum comosum do not respond to variations of red light, although sensitivity to blue light pathway is retained (Roelfsema et al. 2006). In the green orchid Paphiopedilum harrisianum , whose stomata are devoid of green plastids plast, only retains sensitivity to blue light, plus a newly unravelled phytochrome- mediated pathway (Talbott et al. 2002). Here, we studied mainly response to blue and red light together, but previous measures showed that albino did not respond to red light alone (Fig. 7).

Can these differences have any impact on the drying phenomenon that affects more frequently albinos (Table 2 and Fig. 1)? Leaf necrosis, especially leaf necrosis beginning from the edge, could be attributed to overheating symptoms (Bilger et al. 1984), and leaf temperature is regulated by water evaporation. Stomata opening can affect leaf temperature (Merlot et al. 2002), and mutants impaired in regulation of stomata opening often exhibit thermal differences with wild individuals (Hashimoto et a. 2006). Although overheating may not be linked with a real drought, but more with punctual increase of light and temperature. Overheating symptoms are frequently seen on plants overexposed to light (Bilger et al. 1984): light amount in BOI and MON are rather different (respectively barely over 100 µmol.m -2 .s -1 and 1000 µmol.m -2 .s -1 ), which can explain the high frequency of drying at MON and the absence of drying at BOI

We found little direct evidence of overheating. First, although albinos leaves were hotter than green one at ambient light, albino leaf temperature did not increase faster than green one when artificially increasing PAR. We did not find any evidence of short-term responses to light excess, e.g. energy dissipation through photorespiration and production of zeaxanthine from the xanthophyll cycle (Björkman & Demmig-Adams, 1995; Demmig-Adams & Adams, 1995). However, our experiments were performed early in the morning and not at the highest temperature, where overheating may occur. Thus our conditions may not refect the some conditions that may sometime occur and entail overheating (indeed, they were not performed on leaves undergoing drying). As mentioned above, albino leaves temperature was higher than green one at beginning of our experiments, suggesting that in natural, non-transitory conditions, they may thermoregulate less. Moreover, elimination of excess light by way of photosynthetic apparatus is not possible in albinos that have 100 time less chlorophyll and 14 time less carotenoids: thus, we can suspect that energy dissipation would be lower than in

22 green. Although definite evidence is lacking, we propose that albino bear less well some light or drought conditions, and, due to bad heat regulation, undergo more frequently drying as compared to green individuals.

Albino increased sensitivity to pathogens and herbivores Drying is not the lonely plague affecting albinos: they also harvest a higher quantity of insects and mites, and their leaves and fruits are more often predated than these of green indviduals (Table 4). This can, in part, contribute, together with limited C availability, to reduce fruit size. Several reason can explain the higher load on albinos. First, a higher heterotrophy correlates with a higher N content, since part of the recovered fungal biomass is respired into

CO 2, increasing the C/N ratio: this was indeed observed in our samples (Fig. 9). Richness in N contribute to attract more herbivores (Mattson, 1980; Gange & West, 1994; Kursar et al. 2003), so that albinos may be more attractive that green individuals. Moreover, the thinner cuticle may allow fungi (such as Pseudocercospora sp., a common leaf pathogen; Deighton, 1987) and mites to more efficiently invade or penetrate the plant tissues. In addition, the leaf colour may contribute to attract more animals (humans often selectively harvest or destroy albino orchids, M.-A. Selosse, pers. obs.). Although it is unlikely that albinos are a strong source for co-evolution with animals, some learning process may even allow some animals to prefer albino leaves, since their colour is a signal for more fragile, N-richer leaves (with perhaps more limited defences due to lower basal metabolism). Last, there may be positive interaction between necrosis and biotic aggressions, via a decrease in plant defences abilities (Breshears, 2005, McDowell et al. 2008). Then, cause of shoot decay appears multifactorial and it results in a general fitness decay in June, when fruits are maturing.

Leaves and MH evolution To summarize previous results, albino have a 64 times lower fitness than green in MON and, although the study was less detailed and significance more limited by individual numbers, drying was not observed at BOI and fruits were still there in late June. We identified several possible causes for this, including (i) reduced C resources, (ii) increased herbivory and pathogens load and (iii) a frequent shoot drying supposedly linked to more overheating in albinos. As previously mentionned, the dryer Mediterranean environment and higher light exposure at MON may explain why drying is more frequent than in BOI. Although the exposure to high light levels is not unexpected for partly photosynthetic MX species such as

23 Cephalanthera , this conditions contrasts with that of MH plants that frequently grow in dark forest, under denser canopy and wetter forests (Leake 1994). Thus, the conditions in which albinos settle at MON, near adults that produce the seeds, may be far from optimal for shoot survival – but conditions at BOI may entail fewer damages, as evidenced by the fact that albino keep their leaves (no leaf drying observed, cf Fig. 1 of Julou et al. 2005). Moreover, the drying process is in itself an important stage in fitness reduction of albinos (Table 5 and 7). Leaves seem to play a key role in the drying process, because (1) the drying starts from the leaves (Fig. 1), (2) water exchange are more likely to occur here, through stomata. Actually, MH plants have reduced, scaly leaves, and are usually devoid of stomata (Leake 1994: Roy et al. 2009). This is the most obvious difference between albinos and MH plants: the former still thus pay the cost of leaves and stomata… This difference indeed underlines two ironical issues to the retention of these features: the leaves are of no more use, since light capture is no more required, and stomata, whose main role is to counterbalance limiting concentrations of

CO 2, are no more necessary due to loss of photosynthesis. Moreover, adaptation of stomatal water exchange to environmental fluctuations is likely to be disturbed by the absence of plastids in guard cells, as stated above. Indeed, the loss of regulation by red light cues may be less relevant under the denser green cover at BOI, where light is thus more depleted in red radiations (but assessing this specific point needs comparative study of more population) may be and therefore are absent from MH plants). Anyway, leaves in albinos are unnecessary, yet they are likely to contribute to drying. In addition, they represent carbon sinks, targets for parasites such as Pseudocercopsora sp. or criquets (Table 4) – we also speculated on their role as a cue for high-N, low-defence plants. Thus, in a species that is normally partly photosynthetic (40 to 85 % of heterotrophy for C. damansonium ; Gebauer & Meyer, 2003 ;

Julou et al. 2005), and thus bears leaves to capture light and CO 2, the sudden emergence of an albino phenotype entails problems in albinos and contributes to their low success. Thus, evolution toward MH nutrition may have been selected only in certain habitats, where the cost of non-photosynthetic relict leaves is decreased. In some wet forests, small relict leaves can remain, as in C. austinae (Taylor & Bruns, 1997), or they disappear stepwise as heterotrophy increase. There is evidence that a continuum ranging from leafy, green orchids to smaller- leaved MX orchids, such as Epipactis microphylla (Selosse et al. 2004) or Limodorum abortivum (Girlanda et al. 2006), and to nearly MH species with no leaves such as Corallorhiza trifida (Zimmer et al. 2008). Thus, during the evolution to more MH strategies, there is evidence for co-evolution of leaf size and level of heterotrophy, not only because

24 smaller leaves mean less photosynthetic C, but also because leaf size and stomata number has to decrease simultaneously for hydric reasons. We suggest that part of the lack of success of albinos may result of a violation of this necessary coevolution. An interesting comparison can be drawn with plant parasitic on other plants also displaying a continuum from pure autotrophs to pure heterotrophs (Selosse & Roy, 2009) – again, it rather looks like a continuum from leafy, green plants to partly parasitic plants with smaller leaves, and purely heterotrophic parasitic plants with reduced leaves (Press & Graves 1995). Although in this case, high transpiration rate is necessary achieve sap derivation from the host, and thus that the evolution may differ is some respect of that leading to MH orchids, there is evidence that leaves are no more adaptation in heterotrophic plants.

From mixotrophy to mycoheterotrophy: is albinism on the way? Albinism is a clue to detect MX species, already dependant on their fungi for their carbon nutrition, although other adaptations are required to become a true MH and lose its photosynthesis. Mixotrophy itself is already a great step evolutionary step toward complete MH strategy (Selosse et al. 2004; Bidartondo et al., 2004; Selosse & Roy 2009). However, albinism common only in the orchid tribe Neottieae, and although some MX species show aphyllous variants, albinism has never been discovered in any other green terrestrial species to our knowledge (Zimmer et al. 2007): achlorophyllous mutant often occur (Feldmann, 1991) but do not survive in natura . Thus, albinism may be a specific feature to Neottieae, for unknown reasons, and maybe related a frequent evolution to MH nutrition, as 33% of Neottieae did (Bateman et al. 2005). In other clades, there is no evidence that albinism is necessary, and aphylly may be an alternative pathway to MH evolution – as suggested by the model above, as long as fungal C is available, it may be more easy to lose leaves than only photosynthesis, while retaining white leaves, and albinism may be fit only after leaf reduction or colonization of a more favourable habitat. This suggests that the most important factor to select MH is not the emergence of albinos, but the mixotrophy itself.

Selecting for mixotrophy

If MX is that little costly for common mycorrhizal networks, then this nutrition may be widespread. Actually, MX was discovered in Neottieae but the more species are investigated, the more MX are discovered in forest under-storey, among orchids and Pyroloids, green relatives of MH Monotropoids (Tedersoo et al. 2007; Zimmer et al. 2007). Then, transition

25 toward complete MH may be facilitated in such families. Genomic studies focused on chloroplast evolution would be of high interest to understand MX and MH evolution, and to know if selection pressure on photosynthetis genes is relaxed already in MX, allowing more albinism to appear. At least some MH have chloroplastic genes, as matK and rbcL genes are easily amplified but some do not, underlining a possible chloroplast loss (Cameron, 2004; Barrett & Freudenstein, 2008). Albino do have chloroplast but without chlorophyll, and in lower number than green individuals. Genetic studies would be needed to understand if albinism is genetically determined and if mutations occur on chloroplast genome or nuclear genome. Then, comparison of green and albino chloroplast genome would help detecting selection or relaxed selection over photosynthetic genes. It could help understanding why some MH lineages still produce some chlorophyll or keep rbcL gene. Studying chloroplast evolution in MH would also permit to compare it with parasitic plants history, whose chloroplast has been reduced and rearranged in many ways, but also showed strong convergences (Krause, 2008).

Outline

These albinos leaded to investigate green individuals carbon budget and its variations. Heterotrophy seems to vary over the season, and green orchid rely on fungal carbon mainly for growth and underground stages. The photosynthetic carbon is used later, mainly for fruits. Then it underlined that albino may lack some carbon at fruiting time, as thay do not compensate with photosynthesis. Not only their carbon budget is reduced, but their leaf physiology and their bright colour make them more sensitive to light stresses and herbivores. Then, albinism seems to be counter-selected in dry and open area due to the higher cost of non-photosynthetic leaves in such environment. Nevertheless, this trait can be maintained by mutation and by selection as a cryptopolymorphism within mixotrophic species. Albinism may be fixed in more wet and dark environment. Even these albino are counter-selected, they underlined that leaf should be reduced and defences against light-stress and herbivores should be acquired (as pigments) before completely loosing photosynthesis in mycoheterotrophy evolution.

LIST OF FIGURES

26 Figure 1. Albino and green individual of C. damasonium in MON, on 15/06/2007. Albino had already lost its leaves, whereas green individual (here non-flowering) do not present any drying sign.

Figure 2. Phenology of albino (white circles, dotted lines) and green C. damasonium (black circles, plain lines) in MON in 2007 (a) sum of individuals per sampling date and (b) percent of individuals blooming (grey) or fruiting (black) per sampling date and per phenotype (over n=50 albino and n=200 green individuals).

Figure 3. Thin cuttings of albino (a) and green leaves (b) of C. damasonium showing the lack of chloroplast (ch) in albino leaves, or their lower concentration (v : vasculary bundle) and focus on stomata (s) on albino leaves (a) and on green ones (b). Chloroplast are present in both guard cells (g) but are yellow in albino whereas there are more coloured in green cells. Cuticle (c) of albino are two times thinner than green one.

Figure 4. Punctual measures of gas and water exchanges in MON (2006-2008).

Figure 5. Reaction to PAR increase of CO2 assimilation, stomata conductance and difference of temperature between air and leaf temperature in MON (2008).

Figure 6. Reaction to PAR increase of CO2 assimilation, stomata conductance and difference of temperature between air and leaf temperature in BOI (2006).

Figure 7. Green (black circle) and albino (white circle) response to PAR variations when PAR is composed of red light only, at MON, in 2007.

Figure 8. 13C (a) and 15N (b) of green C. damasonium between march and June 2007 in MON, compared to autotroph and mycoheterotroph references. Grey triangles: Clematis , Grey squares: Rubia peregrina , black circles: C. damasonium leaves, black diamond: C. damasonium fruits, white diamonds: C. damasonium albino fruits. Grey line: C. damasonium albino shoot and leaves on 26/06/2007 and black line: Inocybe sp. basidiocarp on 10/11/2007.

Figure 9. (a) Percent of mixotrophy inferred from 13C (a) and 15N (b) signature of C. damasonium leaves (lines) or fruit (signs). Calculation were made either taking autotroph

27 reference of each single date (dotted lines or grey signs) or a mean autotroph reference over the whole spring (plain lines or black signs) and taking as a mycoheterotroph control either albino shoot and leaves (grey lines or circles) or Inocybe sp. basidiocarp (black lines or triangles). (b) C/N (a) and percentage of N in leaves (b) of C. damasonium in 2007 in MON. Grey triangles: Clematis , Grey squares: Rubia peregrina , black circles: C. damasonium leaves, black diamond: C. damasonium fruits, white diamonds: C. damasonium albino fruits. Grey line: C. damasonium albino shoot and leaves on 26/06/2007 and black line: Inocybe sp. basidiocarp on 10/11/2007.

Figure 10. Inter-organs differences of 13C and 15N in C. damasonium green and albino in Montferrier at fruiting time. Different letters show that isotope signature are significantly different based on pairwise Mann-Whitney tests (P<0.05). The first letter correspond to 13C and the second to 15N. Grey circle group measures of different organs form the same species.

Figure 11. Inter-organs differences of 13C and 15N in C. longifolia and E. helleborine in Fontaret at fruiting time. Different letters show that isotope signature are significantly different based on pairwise Mann-Whitney tests (P<0.05). The first letter correspond to 13C and the second to 15N. Grey circle group measures of different organs form the same species.

Figure 12. Inter-organs differences of 13C and 15N in C. damasonium , C. rubra and E. helleborine in Chauriat at fruiting time. Different letters show that isotope signature are significantly different based on pairwise Mann-Whitney tests (P<0.05). The first letter correspond to 13C and the second to 15N. Grey circle group measures of different organs form the same species.

Figure 13. Percentage of heterotrophy calculated on 13C for the four Neottieae species in three sites.

Figure 14. Drying and pathogens symptoms on albino leaves in MON and BOI ; a, Leaf infected by Pseudocercopsora sp. (black spots) without drying in BOI, June 2006 ; b, Leaf infected by pseudocercospora and drying, on 20/06/07 in MON; c, leaf infected by Pseudocercopsora sp. and half dried on 20/06/07 in MON; d, leaf not infected by

28 Pseudocercopsora sp. but entirely dried on 20/06/07 in MON. Complete loss of leaves is reported on Fig. 1.

Figure 15. Fruit size in June of fruits produced after manual crosses. Autogamy refers to flowers enclosed in bags (not manually pollinated), geitonogamy to flowers manually pollinated with it own pollinia, homogamy to flowers manually pollinated with pollinia from another individual of the same type, heterogamy to flowers pollinated with pollinia from a different individual type and not controlled refers to flowers naturally pollinated and enclosed in bag to be protected from herbivores. Albino fruits are represented by white circle, green fruits by black circle, A: albino, G: green individual, ?: not controlled.

Figure 16. Green C. damasonium open flower after 3 days opening, showing pending pollinia.

Figure 17. Scheme of factors affecting albino fitness. Link are reprensented by arrows. If the consequence for albino is negative, a grey circle with a minus sign appears on the arrow. If the consequence for albino is positive, then the grey circle has a plus sign on it. Consequences for green individuals are not shown. As it is a comparison between albino and green, a negative effect on albino can be considered as a negative effect compared to effect for green individuals. Grey squares represent albino individuals, black squares, green individuals. At the bottom of the scheme, long-term consequences are highlighted, and minus signs may result in less new albino individuals whereas plus sign may increase albino number.

29 Table 1. Temporal dynamic of albinos and green individuals in C. damasonium populations between 2006 and 2008 at MON and BOI. Green Difference Albino MON Number % (on %) Number % (1) (1) Number of individuals seen in 2006 99 76.2 > 31 23.8 Number of individuals seen in 2007 423 75.3 > 73 14.7 Number of individuals seen in 2008 449 93.5 > 31 6.5 BOI Number of individuals seen in 2006 6 37.5 < 10 62.5 Number of individuals seen in 2007 10 43.5 < 13 56.5 Number of individuals seen in 2008 10 52.6 > 9 47.4

MON Number % (on %) Number % (2) (3) (2) (3) Individuals seen one year only 591 77.0 < 90 81.1 Individuals seen two consecutive years only 146 19.0 > 8 7.2 Individuals seen over three years 28 3.6 > 3 2.7 Individuals dormant in 2007 2 0.3 < 10 9.0 Total 767 > 111 Mean life span (over three years sampling) 1.26±0.26 > ns 1.28±0.31 BOI Individuals seen one year only 10 55.5 < 12 54.5 Individuals seen two consecutive years only 8 44.5 > 5 22.7 Individuals seen over three years 0 0 < 0 0 Individuals dormant in 2007 0 0 < 5 22.7 Total 18 < 22 Mean life span (over three years sampling) 1.44±0.25 < ns 1.68±0.42

(1) % of the total number of individuals per population (both phenotype) seen the same year. (2) The total can be shared into individuals seen one year, in two consecutive years only and over three years and the remaining ones, dormant in 2007. (3) % of the total number of individuals with same phenotype over the three years sampling.

30 Table 2. Shoot emergence and reproductive cycle of albinos and green individuals in the MON C. damasonium population in 2007 (ns, difference not significant; *, P<0.01; **, P<0.001; ***, P<0.0001 – according to ANOVA).

Green Difference Albino Shoot apparition First apparition of the shoot (=J) 26 th ns 26 th March March Mean date of shoot emergence J+ 46±11 ns J+ 31±8 26/04 11/05 Blooming Mean date of first flower opening J+ 49±6 > ** J+ 43±4 14 th May 8th May Mean age at blooming (days) 24.6±6.9 > * 15.7±7.2 Fruiting Mean date of first fruit formation J+ 58±2 > *** J+ 54±5 Mean age at fruiting (days) 40.8±10.1 > *** 25.0±8.4 Shoot drying First date of shoot drying 24 th April ns 2nd April

Mean date of shoot drying J+ 79±4 ns J+ 69±8 13 th June 3rd June Age at shoot drying (days) 50.0±12.4 ns 45.9±9.1 Percent of individuals drying 17.0 < *** 40.5 Percent of leaf surface dried in late June 11.5 % < *** 76.2 % Percent of individuals completely dried in 7.6 % < *** 51.3 % June

31 Table 3. Comparative morphology and anatomy of albinos and green individuals in the MON and BOI C. damasonium populations (ns, difference not significant; *, P<0.01; **, P<0.001; ***, P<0.0001 – according to ANOVA or Mann-Whitney tests. For pigment detailes compostion, the small number of samples did not allow any statistical test and only the difference sign is indicated).

Diffe- Organ Parameter Year Green Albino rence Shoot Shoot height (cm) MON 2006 24.38±15.4 ns 18.91±3.96 2007 17.20±4.46 > *** 13.17±3.55 2008 19.82±4.53 ns 16.70±2.98 BOI 2007 27.70±8.35 >*** 9.25±3.45 2008 20.80±7.35 ns 27.93±6.52 Shoot number per ns MON 2006 2.3±1.0 1.6±0.6 individual 2007 1.4±0.5 ns 1.24±0.2 2008 1.8 ±0.6 ns 1.5±0.5 Leaf Leaf number MON 2006 3.8±0.5 ns 3.6±0.8 2007 3.2±0.5 ns 3.1±0.4 BOI 2007 3.2±0.3 ns 2.5±0.8 2008 3.3±0.5 ns 3.0±0.3 First leaf (1) surface at MON 2006 7.48±3.7 > *** 4.4±3.2 blooming (cm²) 2007 7.20±2.05 > *** 3.55±1.08 First leaf (1) surface at MON 2007 11.39±2.18 > *** 8.67±0.85 fruiting (cm²) SLA (m²/kg) MON 2007 3.08±0.23 (n=15) < *** 13.43±2.65 (n=7) % of nitrogen in leaves ns MON 2007 3.06±0.20 2.61±0.15 in June Leaf thickness (µm) MON 2006 390±02 (n=5) > *** 220±40 (n=5) Cuticle thickness (µm) MON 2006 12±2 (n=5) > *** 7±1 (n=5) Stomata density (nb per MON 2006 10.9±1.2 (n=10) > *** 9.3±0.8 (n=10) mm²) Chlorophyll MON 2006 3.52±0.86 (n=5) > ** 0.05±0.02 (n=5) concentration (ng/gFW) > ** BOI 2006 3.81±0.75 (n=4) 0.06±0.01 (n=4) (2) MON 2006 2.78±0.56 (n=4) ns 1.57±0.45 (n=2) Chlorophyll a/b BOI 2006 2.37 (n=1) > 1.25 (n=1) Total pigments MON 2006 0.222±0.054 (n=2) > 0.041±0.003 (n=2) (µg/gFW) BOI 2006 0.273 (n=1) > 0.045 (n=1) Total carotenoids MON 2006 0.054±0.024 (n=2) < 0.010±0.005 (n=2) (µg/gFW) BOI 2006 0.042 (n=1) < 0.010 (n=1) % of carotenoids in total MON 2006 33.0±12.2 (n=2) ns 58.9±23.2 (n=2) pigments (weight:weight < BOI 2006 23.3 (n=1) 91.7 (n=1) ) % of light-stress = carotenoids MON 2006 0 0 (Zeaxanthine) BOI 2006 0 = 0 % of root surface ns Root MON 2007 59.5±9.1 (n=4) 63.3 ±10.2 (n=6) colonized when fruiting

(1) The first leave is XXà définirXX. (2) gFW: g of fresh weight

32 Table 4. Observed susceptibility to herbivory and pathogens of albinos and green individuals from MON and BOI C. damasonium populations between 2006 and 2008 (*, P<0.01; **, P<0.001; ***, P<0.0001 – according to ANOVA or Mann-Whitney tests).

Phenotype Target organ Green Difference Albino % of 2007 MON individuals presenting: Herbivory signs (1) Leaf 15.5 <*** 43 Flower or fruit 1.6 <* 5.4 Ants Leaf or shoot 2.6 <* 9.4 Criquet Leaf and flower1 <* 5.4 Caterpillar Leaf and fruit 0 <*** 4 Mites Flower 2 ns 5.4 Spiders ( Tomis ) Flower 2.4 <** 9.5 Aphids Fruit 6 ns 10.8 Any insect (total) Leaf 3.6 <*** 18.8 Fruit 6 <*** 14.8 Any invertebrates (total) Flower 4.4 <*** 14.9

Infection by Pseudocercospora sp. Leaf 35 <*** 55

% of BOI individuals presenting: Herbivory signs in 2007 Leaf 40 ns 15 in 2008 Leaf 10 <* 75 Infection by Pseudocercospora sp. in 2007 Leaf 50 ns 77 in 2008 Leaf 0 <*** 75

33 Table 5. Reproductive apparatus and mating efficiency of albino and green individuals at MON and BOI C. damasonium populations, between 2006 and 2008. Stage Parameter studied Year Green Difference Albino Reproductive % (and number) of MON 2006 3.0% (3) < 6.4% (2) individuals individuals not flowering 2007 21.7% (92) < 27.4% (20) 2008 11.3% (51) > 6.4% (2) BOI 2007 10% (1) < 53.8% (7) 2008 0 = 0 Reproductive height (cm) MON 2008 8.65±2.60 ns 8.00±2.49 Reproductive height (cm) BOI 2008 17.22±4.43 ns 19.85±3.51 Flowers Number of flowers per MON 2006 3.87±1.27 ns 5.00±1.65 shoot 2007 4.40±1.65 ns 3.77±1.45 2008 4.69±1.69 ns 5.93±1.79 BOI 2007 5.0±2.12 >*** 0.91±0.75 2008 3.22±1.5 ns 5.5±2.0 Correlation number of MON 2006 25.4 *** 56.7 *** flowers-shoot size: R² MON 2007 43.1 *** 40.5 *** (%) and significativity MON 2008 48.5 *** 47.1 ** Odour quantity MON 2007 0.32±0.16 ns 0.36±0.18 (µg/flower) Ratio of fruit per flower MON 2007 0.81±0.17 > *** 0.63±0.23 (pollination rate) Fruits Percent of shoots MON 2007 7.6 % < *** 51.3 % disappeared before late June Number of fruits in late MON 2006 2.45±1.35 ns 2.83±1.84 June on remaining shoots 2007 3.53±1.75 > *** 2.65±1.70 2008 3.51±1.45 ns 3.71±1.72 BOI 2007 5.6±1.82 ns 2.25±0.65 2008 2.86±1.30 ns 2.75±1.43 Ratio of fruits per flowers MON 2006 0.61±0.23 ns 0.48±0.23 at individual level on 2007 0.43±0.19 ns 0.38±0.24 remaining shoots MON 2008 0.64±0.25 ns 0.52±0.19 BOI 2007 0.94±0.04 ns 0.92±0.08 2008 0.71±0.18 ns 0.45±0.14 Percentage of fallen fruits MON 2006 39.0±11.5 ns 52.0±11.5 in June (%) calculated on 2007 30.3±16.5 ns 20.2±15.2 remaining shoots 2008 36.9±18.7 ns 47.6±19.1 BOI 2008 29.0±18.0 ns 55.1±14.1 Fruit diameter (cm) (1) MON 2007 0.58±0.09 > ** 0.43±0.08 2008 0.67 ±0.15 > ** 0.35±0.17 BOI 2007 0.52±0.09 ns 0.38±0.03 Fruit length (cm) (1) MON 2007 1.79±0.21 > * 1.55±0.17 2008 1.62±0.38 > ** 1.04±0..47 BOI 2007 2.43±0.27 ns 2.06±0.14 Fruit fresh weight (mg) MON 2008 151.6 > ** 48.3±30.9 (2) ±45.8 Fruit dry weight (mg) (2) MON 2008 64.5±16.6 > ** 28.9±12.9 Percentage of Nitrogen in MON 2007 2.88±0.11 < *** 4.69±0.68 fruits (3) (1) Fruit diameter and length were measured on a subset of individuals in MON (n=30 albino individuals and n=30 green individuals). (2) Fruit fresh and dry weight was measured on n=20 albino and n=20 green fruits. (3) Fruit Nitrogen content was measured on n=5 albino and n=5 green fruits.

34 Other parameters were measured on the whole population.

Table 6. Seeds production and germination for albino and green individuals at MON and BOI C. damasonium populations between 2006 and 2008.

Stage Parameter studied year Green Difference Albino Seeds (2 months Number of seeds per MON 2006 1124±385 > ** 379± 214 old) fruit Mean fertility rate per MON 2006 37.31±7.75 45.68±4.65 fruit (percentage of < * seeds with embryo) Germination Percentage of seeds MON 2006 0.53±0.01 < *** 3.48±0.03 (after 18 months without embryo in buried in soil) (empty testa may have disappeared) Percentage of seeds MON 2006 22.2±10.1 < *** 73.9±12.9 with embryo in contact with fungal mycelium but not deformed Percentage of seeds MON 2006 75.66±10.11 > *** 21.06±13.11 with swollen embryo colonized by fungal mycelium Percentage of MON 2006 0.788±0.018 ns 0.654±0.008 protocorms

Table 7. Comparison of fitness between albino (A) and green individuals (G) in the two populations investigated.

MON BOI Over-year survival G = A G = A Shoot survival during the year G 2x > A G = A Flower number G = A G = A Fruit number in late June G = A G = A Fruit size G2x >A G=A Seeds number G 4x > A ? Seeds viability G 4x > A ? Fitness G 64x > A G = A

35 REFERENCES

Abadie, J. C., U. Puttsepp, G. Gebauer, A. Faccio, P. Bonfante, and M. A. Selosse. 2006. Cephalanthera longifolia (Neottieae, Orchidaceae) is mixotrophic: a comparative study between green and nonphotosynthetic individuals. Canadian Journal of Botany-Revue Canadienne De Botanique 84 :1462- 1477.

Altschul, S. F., T. L. Madden, A. A. Schaffer, J. H. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research 25 :3389-3402.

Amthor, J. S. 1995. Higher Plant Respiration and Its Relationships to Photosynthesis. Pages 71-101 in E.-D. Schulze and B. Caldwell, editors. Ecophysiology of Photosynthesis. Springer, New York.

Archibald, J. M. 2009. The Puzzle of Plastid Evolution. Current Biology 19 :R81-R88.

Badeck, F. W., G. Tcherkez, S. Nogues, C. Piel, and J. Ghashghaie. 2005. Post-photo synthetic fractionation of stable carbon isotopes between plant organs - a widespread phenomenon. Rapid Communications in Mass Spectrometry 19 :1381-1391.

Barkman, T. J., J. R. McNeal, S. H. Lim, G. Coat, H. B. Croom, N. D. Young, and C. W. dePamphilis. 2007. Mitochondrial DNA suggests at least 11 origins of parasitism in angiosperms and reveals genomic chimerism in parasitic plants. Bmc Evolutionary Biology 7:-.

Barrett, C. F., and J. V. Freudenstein. 2008. Molecular evolution of rbcL in the mycoheterotrophic coralroot orchids (Corallorhiza Gagnebin, Orchidaceae). Molecular Phylogenetics and Evolution 47 :665-679.

Bateman, R. M., P. M. Hollingsworth, J. Squirell, and M. L. Hollingsworth. 2005. Phylogenetics: Neottieae. in P. A.M., P. J. Cribb, M. W. Chase, and F. N. Rasmussen, editors. Genera Orchidacearum (4): Epidendroideae. Oxford University Press, Oxford, UK.

Berger, D., and T. Altmann. 2000. A subtilisin-like serine protease involved in the regulation of stomatal density and distribution in Arabidopsis thaliana. Genes & Development 14 :1119-1131.

Beyrle, H. F., and S. E. Smith. 1993. Excessive carbon prevents greening of leaves in mycorrhizal seedlings of the terrestrial orchid Orchis morio . Lindleyana 8:97-99.

Bidartondo, M. I., B. Burghardt, G. Gebauer, T. D. Bruns, and D. J. Read. 2004. Changing partners in the dark: isotopic and molecular evidence of ectomycorrhizal liaisons between forest orchids and trees. Proceedings of the Royal Society of London Series B-Biological Sciences 271 :1799-1806.

Bilger, H. W., U. Schreiber, and O. L. Lange. 1984. Determination of Leaf Heat-Resistance - Comparative Investigation of Chlorophyll Fluorescence Changes and Tissue Necrosis Methods. Oecologia 63 :256- 262.

Birky, C. W. 1995. Uniparental Inheritance of Mitochondrial and Chloroplast Genes - Mechanisms and Evolution. Proceedings of the National Academy of Sciences of the United States of America 92 :11331-11338.

Björkman, O., and B. Demmin-Adams. 1995. Regulation of Photosynthesis Light Energy Capture, Conversion, and Dissipation in Leaves of Higher Plants. Pages 17-47 in E.-D. Schulze and B. Caldwell, editors. Ecophysiology of Photosynthesis. Springer, New York.

Breshears, D. D., N. S. Cobb, P. M. Rich, K. P. Price, C. D. Allen, R. G. Balice, W. H. Romme, J. H. Kastens, M. L. Floyd, J. Belnap, J. J. Anderson, O. B. Myers, and C. W. Meyer. 2005. Regional vegetation die- off in response to global-change-type drought. Proceedings of the National Academy of Sciences of the United States of America 102 :15144-15148.

36 Brzosko, E. 2002. The dynamics of Listera ovata populations on mineral islands in the Biebrza national park. Acta Societatis Botanicorum Poloniae 71 :243-251.

Cameron, K. M. 2004. Utility of plastid psaB gene sequences for investigating intrafamilial relationships within Orchidaceae. Molecular Phylogenetics and Evolution 31 :1157-1180.

Damesin, C., and C. Lelarge. 2003. Carbon isotope composition of current-year shoots from Fagus sylvatica in relation to growth, respiration and use of reserves. Plant Cell and Environment 26 :207-219.

Damesin, C., S. Rambal, and R. Joffre. 1998. Seasonal and annual changes in leaf delta C-13 in two co-occurring Mediterranean oaks: relations to leaf growth and drought progression. Functional Ecology 12 :778-785.

Deighton, F. C. 1987. New Species of Pseudocercospora and Mycovellosiella, and New Combinations into Pseudocercospora and Phaeoramularia. Transactions of the British Mycological Society 88 :365-391.

Delforge, P. 1998. Des Epipactis helleborine (L.) CRANTZ dépourvus de chlorophylle dans les environs de Bruxelles. Naturalistes Belges 79 :124-130.

Demmig-adams, B., and W. W. Adams. 1992. Photoprotection and Other Responses of Plants to High Light Stress. Annual Review of Plant Physiology and Plant Molecular Biology 43 :599-626.

Feldmann, K. A. 1991. T-DNA Insertion Mutagenesis in Arabidopsis - Mutational Spectrum. Plant Journal 1:71- 82.

Gange, A. C., and H. M. West. 1994. Interactions between Arbuscular Mycorrhizal Fungi and Foliar-Feeding Insects in Plantago-Lanceolata L. New Phytologist 128 :79-87.

Gebauer, G., and M. Meyer. 2003. N-15 and C-13 natural abundance of autotrophic and mycoheterotrophic orchids provides insight into nitrogen and carbon gain from fungal association. New Phytologist 160 :209-223.

Girlanda, M., M. A. Selosse, D. Cafasso, F. Brilli, S. Delfine, R. Fabbian, S. Ghignone, P. Pinelli, R. Segreto, F. Loreto, S. Cozzolino, and S. Perotto. 2006. Inefficient photosynthesis in the Mediterranean orchid Limodorum abortivum is mirrored by specific association to ectomycorrhizal Russulaceae. Molecular Ecology 15 :491-504.

Hashimoto, M., J. Negi, J. Young, M. Israelsson, J. I. Schroeder, and K. Iba. 2006. Arabidopsis HT1 kinase controls stomatal movements in response to CO2. Nature Cell Biology 8:391-U352.

Julou, T., B. Burghardt, G. Gebauer, D. Berveiller, C. Damesin, and M. A. Selosse. 2005. Mixotrophy in orchids: insights from a comparative study of green individuals and nonphotosynthetic individuals of Cephalanthera damasonium . New Phytologist 166 :639-653.

Kerstiens, G. 1996. Signalling across the divide: A wider perspective of cuticular structure-function relationships. Trends in Plant Science 1:125-129.

Krause, K. 2008. From chloroplasts to "cryptic" plastids: evolution of plastid genomes in parasitic plants. Current Genetics 54 :111-121.

Kursar, T. A., and P. D. Coley. 2003. Convergence in defense syndromes of young leaves in tropical rainforests. Biochemical Systematics and Ecology 31 :929-949.

Leake, J. R. 1994. The Biology of Myco-Heterotrophic (Saprophytic) Plants. New Phytologist 127 :171-216.

Lopez, S., F. Rousset, F. H. Shaw, R. G. Shaw, and O. Ronce. 2008. Migration load in plants: role of pollen and seed dispersal in heterogeneous landscapes. Journal of Evolutionary Biology 21 :294-309.

Mairold, F., and F. Weber. 1950. Notiz über Cephalanthera albinos. Protoplasma 39 :275-277.

37 Mattson, W. J. 1980. Herbivory in Relation to Plant Nitrogen-Content. Annual Review of Ecology and Systematics 11 :119-161.

Maunoury, F., D. Berveiller, C. Lelarge, J. Y. Pontailler, L. Vanbostal, and C. Damesin. 2007. Seasonal, daily and diurnal variations in the stable carbon isotope composition of carbon dioxide respired by tree trunks in a deciduous oak forest. Oecologia 151 :268-279.

McDowell, N., W. T. Pockman, C. D. Allen, D. D. Breshears, N. Cobb, T. Kolb, J. Plaut, J. Sperry, A. West, D. G. Williams, and E. A. Yepez. 2008. Mechanisms of plant survival and mortality during drought: why do some plants survive while others succumb to drought? New Phytologist 178 :719-739.

Molvray, M., P. J. Kores, and M. W. Chase. 2000. Polyphyly of mycoheterotrophic orchids and functionnal influences on floral and molecular characters. Pages 441-447 in K. L. Wilson and D. A. Morrison, editors. Monocots: systematics and evolution. CSIRO, Melbourne.

Phillips, D. L., and J. W. Gregg. 2001. Uncertainty in source partitioning using stable isotopes (vol 172, pg 171, 2001). Oecologia 128 :304-304.

Press, M. C., and J. D. Graves. 1995. Parasitic Plants. Chapmann & Hall.

Proffit, M., B. Schatz, J. M. Bessiere, C. Chen, C. Soler, and M. Hossaert-McKey. 2008. Signalling receptivity: Comparison of the emission of volatile compounds by figs of Ficus hispida before, during and after the phase of receptivity to pollinators. Symbiosis 45 :15-24.

Rasmussen, H. N. 1995. Terrestrial Orchids – From Seed to Mycotrophic Plant. Cambridge University Press.

Rasmussen, H. N. 2002. Recent developments in the study of orchid mycorrhiza. Plant and Soil 244 :149-163.

Reyes-Prieto, A., A. Moustafa, and D. Bhattacharya. 2008. Multiple genes of apparent algal origin suggest ciliates may once have been photosynthetic. Current Biology 18 :956-962.

Robert, C., M. O. Bancal, B. Ney, and C. Lannou. 2005. Wheat leaf photosynthesis loss due to leaf rust, with respect to lesion development and leaf nitrogen status. New Phytologist 165 :227-241.

Roelfsema, M. R. G., K. R. Konrad, H. Marten, G. K. Psaras, W. Hartung, and R. Hedrich. 2006. Guard cells in albino leaf patches do not respond to photosynthetically active radiation, but are sensitive to blue light, CO2 and abscisic acid. Plant Cell and Environment 29 :1595-1605.

Rosati, G., L. Barsanti, V. Passarelli, A. Giambelluca, and P. Gualtieri. 1996. Ultrastructure of a novel nonphotosynthetic Euglena mutant. Micron 27 :367-373.

Roy, M., T. Yagame, M. Yamato, K. Iwase, C. Heinz, A. Faccio, P. Bonfante, and M. A. Selosse. 2009. Ectomycorrhizal Inocybe species associate with the mycoheterotrophic orchid Epipogium aphyllum but not with its asexual propagules. Annals of Botany in press .

Salmia, A. 1986. Chlorophyll-free form of Epipactis helleborine (Orchidaceae) in SE Finland. Annals of Botany Fennici 23 :49-57.

Salmia, A. 1989a. Features of endomycorrhizal infection of chlorophyll-free and green forms of Epipatis helleborine (Orchidaceae). Annals of Botany Fennici 26 :15-26.

Salmia, A. 1989b. General morphorlogy and anatomy of chlorophyll-free and green forms of Epipactis helleborine (Orhcidaceae). Annals of Botany Fennici 26 :95-105.

Saldarriaga, J. F., F. J. R. Taylor, P. J. Keeling, and T. Cavalier-Smith. 2001. Dinoflagellate nuclear SSU rRNA phylogeny suggests multiple plastid losses and replacements. Journal of Molecular Evolution 53 :204- 213.

Scacchi, R., G. Deangelis, and P. Lanzara. 1990. Allozyme Variation among and within 11 Orchis Species (Fam Orchidaceae), with Special Reference to Hybridizing Aptitude. Genetica 81 :143-150.

38 Scappaticci, C., and G. Scappaticci. 1998. Epipactis micorphylla (EHRHARDT)SWARTZ lusus rosea SOO. in Cahiers de la Socitété Française d'Orchidophilie, 1eres journées Rencontres Orchidophiles Rhônes- Alpes. Lyon.

Selosse, M. A., A. Faccio, G. Scappaticci, and P. Bonfante. 2004. Chlorophyllous and achlorophyllous specimens of Epipactis microphylla (Neottieae, Orchidaceae) are associated with ectomycorrhizal septomycetes, including truffles. Microbial Ecology 47 :416-426.

Selosse, M. A., F. Richard, X. H. He, and S. W. Simard. 2006. Mycorrhizal networks: des liaisons dangereuses? Trends in Ecology & Evolution 21 :621-628.

Selosse, M. A., and M. Roy. 2009. Green plants that feed on fungi: facts and questions about mixotrophy. Trends in Plant Sciences in press .

Selosse, M. A., M. Weiss, J. L. Jany, and A. Tillier. 2002. Communities and populations of sebacinoid basidiomycetes associated with the achlorophyllous orchid Neottia nidus-avis (L.) LCM Rich. and neighbouring tree ectomycorrhizae. Molecular Ecology 11 :1831-1844.

Shefferson, R. P., T. Kull, and K. Tali. 2005. Adult whole-plant dormancy induced by stress in long-lived orchids. Ecology 86 :3099-3104.

Shefferson, R. P., J. Proper, S. R. Beissinger, and E. L. Simms. 2003. Life history trade-offs in a rare orchid: The costs of flowering, dormancy, and sprouting. Ecology 84 :1199-1206.

Shimazaki, K. I., M. Doi, S. M. Assmann, and T. Kinoshita. 2007. Light regulation of stomatal movement. Annual Review of Plant Biology 58 :219-247.

Smedley, M. P., T. E. Dawson, J. P. Comstock, L. A. Donovan, D. E. Sherrill, C. S. Cook, and J. R. Ehleringer. 1991. Seasonal Carbon Isotope Discrimination in a Grassland Community. Oecologia 85 :314-320.

Talbott, L. D., J. X. Zhu, S. W. Han, and E. Zeiger. 2002. Phytochrome and blue light-mediated stomatal opening in the orchid, Paphiopedilum. Plant and Cell Physiology 43 :639-646.

Taylor, D. L., and T. D. Bruns. 1997. Independent, specialized invasions of ectomycorrhizal mutualism by two nonphotosynthetic orchids. Proceedings of the National Academy of Sciences of the United States of America 94 :4510-4515.

Tedersoo, L., P. Pellet, U. Koljalg, and M. A. Selosse. 2007. Parallel evolutionary paths to mycoheterotrophy in understorey Ericaceae and Orchidaceae: ecological evidence for mixotrophy in Pyroleae. Oecologia 151 :206-217.

Tuba, Z., H. K. Lichtenthaler, Z. Csintalan, Z. Nagy, and K. Szente. 1996. Loss of chlorophylls, cessation of photosynthetic CO2 assimilation and respiration in the poikilochlorophyllous plant Xerophyta scabrida during desiccation. Physiologia Plantarum 96 :383-388.

Van den Kinderen, G. 1995. A method for the study of field germinated seeds of teerstrial orchids. Lindleyana 10 :68-73.

Williams, J. H. H., and J. F. Farrar. 1990. Control of Barley Root Respiration. Physiologia Plantarum 79 :259- 266.

Zimmer, K., C. Meyer, and G. Gebauer. 2008. The ectomycorrhizal specialist orchid Corallorhiza trifida is a partial myco-heterotroph. New Phytologist 178 :395-400.

Zimmer, K., N. A. Hynson, G. Gebauer, E. B. Allen, M. F. Allen, and D. J. Read. 2007. Wide geographical and ecological distribution of nitrogen and carbon gains from fungi in pyroloids and monotropoids (Ericaceae) and in orchids. New Phytologist 175 :166-175.

39 1 cm

Figure 1 350 a 300

250

200

150

100

0 19/03/2007 08/04/2007 28/04/2007 18/05/2007 07/06/2007 27/06/2007 17/07/2007 Sum of individuals per date 0.45 b 400.4 0.35 300.3 0.25 200.2 0.15 100.1 0.05 0 19/03/2007 08/04/2007 28/04/2007 18/05/2007 07/06/2007 27/06/2007 17/07/2007 Percent of individuals blooming or fruiting per date for each type

growth flower fruit

Figure 2 v v

0.8 mm 0.8 mm

s s c c

g g ch ch g g

100 µm 100 µm

Figure 3 ) -1

.s 4 -2

.m 2 2 0 -2 (µmol CO Net photosynthesis Net -4 PAR=0 PAR=100 PAR=0 PAR=100 PAR=0 PAR=100 PAR=0 PAR=100 PAR=0 PAR=100 June2006 May 2006 June2007 June2008 June 2006 Boigneville

0.15

) 0.12 -1 .s

-2 0.09 0.06 O m O 2 0.03 0

(mol H -0.03

Stomatal conducatnce -0.06 PAR=0 PAR=100 PAR=0 PAR=100 PAR=0 PAR=100 PAR=0 PAR=100 PAR=0 PAR=100 June2006 May 2006 June2007 June2008 June 2006 Boigneville

Figure 4 8 7 ) 6 -1

.s 5 -2 4 .m

2 3 2 1 0 -1 0 200 400 600 800 1000 1200 1400 1600 1800 2000 (µmol CO Net photosynthesis -2 -3

PAR 0.14 ) -1

.s 0.12 -2 0.1 O O m

2 0.08

0.06

0.04 (mol (mol H

Stomatal conducatnce 0.02

0 0 200 400 600 800 1000 1200 1400 1600 1800 2000

1.5 PAR 1

0.5

-0.5

-1

-1.5 0 200 400 600 800 1000 1200 1400 1600 1800 2000 -2 PAR (µmol m -2 .s -1 )

Leaf temperature –air temperature °C Figure 5 6

4 ) -1

.s 3 -2 2 .m 2 1

0 0 200 400 600 800 1000 1200 1400 1600 1800 2000 -1 (µmol CO Net photosynthesis 0.1

0.08

0.06 ) -1

.s 0.04 -2

0.02 O O m 2

0 0 200 400 600 800 1000 1200 1400 1600 1800 2000

(mol (mol H -0.02 Stomatal conducatnce

-0.04

-0.06 1

-1

-2

-3

-4

-5 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Leaf temperature –air temperature °C Figure 6 PAR (µmol m -2 .s -1 ) 0.12 )

-1 0.1 .s -2

0.08 O O m 2

0.06 (mol (mol H

Stomatal conducatnce 0.04

0.02

0 0 200 400 600 800 1000 1200 1400 1600

PAR (µmol m -2 .s -1 )

Figure 7 13 Figure 8 Figure 15 C (‰)

N (‰) 2 07/06/2007 18/05/2007 28/04/2007 08/04/2007 19/03/2007 -32 -30 -28 -26 -24 -22 -20 90/070/420 80/071/520 70/072 07/06/2007 18/05/2007 28/04/2007 08/04/2007 19/03/2007 10 15 20 25 -5 0 5 growth growth flower flower C. damasonium C. green R.peregrina fruit fruit /620 17/07/2007 7/06/2007 /620 17/07/2007 7/06/2007 green fruit green liofruitalbino liofruitalbino green fruit green C. vitalba C. C. damasonium C. green % of fungal Carbon in C. a c b damasonium green leaves

Figure 9 Figure C/N 90/070/420 80/071/520 70/072 07/06/2007 18/05/2007 28/04/2007 08/04/2007 19/03/2007 100 90/070/420 80/071/520 70/072 07/06/2007 18/05/2007 28/04/2007 08/04/2007 19/03/2007 10 15 20 25 30 35 40 10 20 30 40 50 60 70 80 90 0 5 0 growth C. damasonium C. flower R.peregrina lioleafalbino fruit /620 17/07/2007 7/06/2007 /620 17/07/2007 7/06/2007 C.d green fruit green C.d C. damasonium C. vitalba C. . liofruit albino C.d green 25 Cephalanthera damasonium green

fruit (f,d) 20

15 Inocybe cf. dulcamara (e,c)

leaf (c,b)

N (‰) N 10 15 fruit (d,b) leaf (b,b) Cephalanthera damasonium albino 5

Clematis vitalba (a,a) -5 -30 -28 -26 -24 -22 -20 13 C (‰)

Figure 10 20 Epipactis helleborine Leaves (c,i) Root (g,i) Rhizome (g,i) 15

N (‰) N Fruit (f,d) Shoot (g,h) 15 10

Neottia nidus-avis (e,e) Cephalanthera longifolia 5 Leaves (c,d) Root (d,d) Trifolium arvense (a,a) Rhizome (d,d) 0 Fruit (c,b) Shoot (c,b) Hedera helix (b,a) -5 -32 -30 -28 -26 -24 -22 -20 13 C (‰)

Figure 11 10 Monotropa hypopithys (e,g) Neottia nidus-avis (i,g) 8 Epipactis helleborine Fruit (d,f) 6 Cephalanthera rubra Leaves (d,e) Rhizome (d,d) 4 Shoot (d,d) Flower (g,d) Root (e,d) Shoot (e,d) Fruit (d,d) Rhizome (g,d) 2 Root (h,d) Leaves (d,d) N (‰) N

15 Root (g,d) 0 Rhizome (f,d) Shoot (e,c) Cephalanthera damasonium -2 Vincetoxicum sp. (c,b) -4 Hedera helix (b,b) Cornus sanguineus (a,a) -6

-8 -34 -32 -30 -28 -26 -24 -22 -20 13 C (‰)

Figure 12 % of heterotrophy 200 b 180

160 a 140

120 a a a 100 a a,c 80 a c a a b b a c c b 60 a a b b a b 40 c c

0 L Fr L Fr R Rh S L Fr R Rh S L Fr R Rh S L Fr R Rh S L Fr R Rh S Fl

C. damasonium C. longifolia E. helleborine E. helleborine C. damasonium C. rubra Montferrier Fontaret Fontaret Chauriat Chauriat Chauriat

Figure 13 &

+=(2

(

+=(2 )

+=(2 Figure 14 1 0.9 0.8 Vx? 0.7 VxV 0.6 VxV 0.5 VxA 0.4 Ax? 0.3 AxA AxV 0.2 AxA 0.1 AxA VxV 0 Auto Geitono Homo Hetero Natural gamy gamy gamy gamy crossing

Figure 15 *_/=(2

Figure 16 Green indivdisuals Albino indivdisuals Same mycorrhization Same mycorrhization photosynthesis No photosynthesis

C Budget C Budget pigments

stomata density Shoot production cuticle lower metabolism

Shoot size stomata conductance

odour Number of flowers Defense against Defense against pathogens light stress and herbivores

pollination Number of putative fruits Shoot survival at fruiting time

Green fruits pollinated Albino fruits by albino : pollinated male fitness

Seeds set Seeds set : female fitness Underground survival

germination dormancy underground germination mortality Long-term consequences

New individuals Same albino New individuals green individual albino or green Figure 17 79.(1*= 79.(1*= = = _=_= `= _= `= _= `= _= `=_=: _= _= *5-&1&39-*7&= *=.,:& = 7*).8(4;*7*)a= 3*<= .38.,-98= .3= 9-*= 9&=4342>`= -&'.9&9= 7*6:.7*2*398= &3)= '7**).3,= 8>89*2=4+=&=7&7*=2>(4-*9*749745-.(=47(-.)_ = :'2.99*)=94= 47).(=4:73&1=4+=49&3>_= = _=*=.,:& =*89=:3*=47(-.)S*=2>(4-S9S749745-*8=7&7*=*9=)S(7.9*=57.3(.5&1*2*39=&:=&48_= &=)S(4:;*79*=*3=-&1&3)*=34:8=&=5*72.8=)S9:).*7=8*8=5&79*3&.7*8=2>(47-.?.*38= (+_ =&79.(1*= `= )*8= -S1S5-47&(S*8= *3= ,S3S7&1= 6:4.6:*= )&:97*8= (-&25.,3438= *(942>(47-.?.*38= &.*39= S9S=.)*39.+.S8`=)439=*11*=7*O4.9=):=(&7'43*=*9=)*=1&?49*=47,&3.6:*_=/*8=7S8:19&98=5*72*99*39=)*= 2.*:==(4257*3)7*=8&='.414,.*=*9=8*8=*=.,*3(*8=S(414,.6:*8_=&38=(*=)*:=.T2*=&79.(1*`=34:8= &;438=S9:).S=51:8=57S(.8S2*39=8&=7*574):(9.43=*9=1*8=(&7&(9S7.89.6:*8=)*=843=*3;.7433*2*39_= *8= 2>(4-S9S749745-*8= 439= 84:;*39= :3*= +147&.843= (4:79*= *9= ).8(7T9*`= 5&7+4.8= 2U2*= 84:9*77&.3*=(422*=(-*?= 5.54,.:2=&5->11:2 `=*3=(438S6:*3(*`=1*:7=7S,.2*=7*574):(9*:7=*89= 84:;*39=2S(433:_=3=349*=5&7=*=*251*=(-*?= _=&5->11:2 =:3=+&.'1*=8:((T8=7*574):(9*:7`=*9= )*8= +7:.98= 6:.= 942'*39_= &= +147&.843= 84:9*77&.3*= 54:77&.9= 1&.88*7= 8:5548*7= :3*= 9*3)&3(*= A= 1&:94,&2.*`= 2&.8= 1*= 5&7+:2= ;&3.11S= &88*?= +479= *9= 1*8= 4'8*7;&9.438= )*= 5411.3.8&9*:78= )S24397*39=1*=(4397&.7*_=/*5*3)&39`=1&=7*574):(9.43=)*=(*99*=*85T(*=7*548*=)&;&39&,*=8:7=1&= 7*574):(9.43=&8*=:S* =;.&= 1*8=':1'.11*8`=*9=)&.11*:78`=1*8=9.,*8=+1*:7.*8=5*:;*39=3*=5&8=U97*=;:*8= 5*3)&39=)*8=).?&.3*8=)&33S*`=(*=6:.=24397*=6:*=(*99*=2>(4-S9S749745-*8=;.9=*9=8*=7*574):.9= *88*39.*11*2*39=84:8=9*77*_=41;7&>= *9=&1_` =,***=8:,,T7*=6:*=1&:94,&2.*=+&.9=5&79.*=)*8=97&.98= )*8=2>(4-S9S749745-*8=2&.8= _=&5->11:2 =)S24397*=1*=(4397&.7*_=&=7*)S(4:;*79*=)*= _=*=.,:& = S9&.9=)43(=14((&8.43=)S9:).*7=51:8=57S(.8S2*39=843=7S,.2*=)*=7*574):(9.43_= 71= 7*88479= 6:*= _= *=.,:& =*89=&114,&2*`=2&.8=3*=574):.9=5&8=)*=3*(9&7`=.1=8&,.9= ):3*= 47(-.)S*= 97.(-*:8*= ):= 54.39= )*= ;:*= )*= 1&= 5411.3.8&9.43_= 7T8= 5*:= )*= 8*8= +7:.98= &77.;*39= A= 2&9:7.9S`=*9=(-&6:*=5.*)=574):.9=+=A=,=+1*:78=*3=24>*33*`='.*3=24.38=6:*= _=&:89.3&* `=:3*= &:97*= 2>(4-S9S749745-*_= 4:7= :3*= 47(-.)S*= 5&7&8.9*`= 43= &:7&.9= 5:= 8&99*3)7*= A= :3= .3;*89.88*2*39= 57S+S7*39.*1= )&38= 1&= 7*574):(9.43`= (*= 6:.= 3*89= 5&8= 1*= (&8= .(.=a= (*99*= 2>(4-S9S749745-*=574):.9=)*8=+*:.11*8=343=(-14745->11.*33*8=*9=5*:=)*=+1*:78_=:.*3=6:*=1&= 9&.11*=)*8=5.*)8=+1*:7.8=84.9=+&.'1*=&:94:7=)*=+*=(2`=43=5*:9=8*=)*2&3)*7=6:*1=*89=1*=8:((T8=)*= 8&= ).85*78.43= *9= )*= 8&= ,*72.3&9.43= 6:.= 54:77&.9= *=51.6:*7= 8&= 7*1&9.;*= S9*3):*= &:= 3.;*&:= ,S4,7&5-.6:*= *9= 1&= )*38.9S= )*= 8*8= 545:1&9.438`= 349&22*39= 7.(-*= ).3).;.):8= )472&398= 4'8*7;S8=1478=)*8=57S1T;*2*398_= _= *=.,:& = 845548*= )43(= &:== ->549-T8*8= )*= 8S1*(9.43= )*= 1&:94,&2.*= (-*?= 1*8= 2>(4-S9S749745-*8_= /*5*3)&39`= (*99*= *85T(*= &= 574'&'1*2*39= +.=S= 51:8= 7S(*22*39= 1&= 2>(4-S9S749745-.*= 6:*= 1*8= &:97*8= *85T(*8= 2>(4-S9S749745-*8=a= *11*= 57S8*39*= *3(47*= )*8= +*:.11*8= )S;*1455S*8= *9= 8&38= 5.,2*39&9.43= &((*884.7*`= 8*8= ,T3*8= (-147451&89.6:*8= 8439= &251.+.&'1*8= *9= 3*= 57S8*39*39= 5&8= )*= 24).+.(&9.43= 2&/*:7*_= 71= *89= &1478= 5488.'1*= 6:*= 1&:94,&2.*=3&.9=5&8=*3(47*=S9S=8S1*(9.433S*=(-*?=(*99*=*85T(*_= Nordic Journal of Botany 27: 460Á468, 2009 doi: 10.1111/j.1756-1051.2009.00465.x, # The Authors. Journal compilation # Nordic Journal of Botany 2009 Subject Editor: Petra Korall. Accepted 14 July 2009

Cephalanthera exigua rediscovered: new insights in the taxonomy, habitat requirements and breeding system of a rare mycoheterotrophic orchid

Henrik Æ. Pedersen, Santi Watthana, Me´lanie Roy, Somran Suddee and Marc-Andre´ Selosse

H. Æ. Pedersen ([email protected]), Botanical Garden and Museum, Natural History Museum of Denmark, Univ. of Copenhagen, Gothersgade 130, DKÁ1123 Copenhagen K, Denmark. Á S. Watthana, Queen Sirikit Botanic Garden, PO Box 7, Mae Rim, Chiang Mai 50180, Thailand. Á M. Roy and M.-A. Selosse, Centre d’Ecologie Fonctionnelle et Evolutive, 1919 route de Mende, FRÁ34293 Montpellier cedex 05, France. Á S. Suddee, Forest Herbarium (BKF), National Park, Wildlife and Plant Conservation Department, Chatuchak, Bangkok 10900, Thailand.

Five mycoheterotrophic species of Cephalanthera have been proposed from tropical Asia. Until recently, all of them were only known from the type specimens, for which reason it has been difficult to judge if some of them were actually conspecific. The recent discovery of two large populations of C. exigua in Thailand made it possible to describe the range of variation of this species in much more detail, and we then found it to be well distinguished from the others. A phylogenetic analysis based on nuclear rDNA ITS sequences corroborated the systematic position of C. exigua at the generic level. Examination of the previously neglected type of C. pusilla from Myanmar lead to the conclusion that this taxon of disputed generic affinity likewise belongs to Cephalanthera, and that it is a clearly distinct species. Finally, we accept that also C. ericiflora from Laos as well as C. gracilis and C. calcarata from Yunnan are probably distinct species. Cephalanthera exigua grows in hill evergreen forest, but we found little floristic similarity between its two localities in Thailand (ISS 22.5%). Ectomycorrhizal trees of the families Fagaceae and Ulmaceae are probably the ultimate carbon source for C. exigua at both sites, as the mycorrhizal partners of this orchid (basidiomycetes of the family Thelephoraceae) obtain their carbon nutrition through ectomycorrhizal association with forest trees. Pollination experiments demonstrated that the nectarless flowers of C. exigua are not spontaneously autogamous. Based on the pollination syndrome and previously published observations on other species of the genus, we suggest that C. exigua has an insect-operated breeding system acting by deceit, and that the pollinators are probably solitary bees.

Orchids without chlorophyll are unable to photosynthesize, Rich., it has recently been demonstrated that individuals and those with strongly reduced levels of chlorophyll can with normal green leaves do not only obtain carbon only perform photosynthesis far below the compensation through photosynthesis, but also to a high extent from point. In both cases, the plants are therefore completely their fungal symbionts; a strategy known as mixotrophy dependent on their mycorrhizae as a carbon source, even at (Gebauer and Meyer 2003, Julou et al. 2005, Abadie et al. their adult stages (Rasmussen 1995), and they are referred 2006). This obviously makes species of Cephalanthera to as being mycoheterotrophic or holomycotrophic (Leake predisposed for developing a mycoheterotrophic strategy, 1994). Orchid species with a permanently mycohetero- and indeed chlorophyll-deficient to chlorophyll-free in- trophic way of life have evolved on several occasions in three dividuals have been observed in each of the normally of the five subfamilies of the Orchidaceae (Molvray et al. mixotrophic species C. damasonium (Renner 1938, 1943, 2000). Therefore, it is hardly surprising that the limit Mairold and Weber 1950, Kohns and Schneider 1993, between autotrophy and mycoheterotrophy is indistinct in Julou et al. 2005), C. longifolia (Renner 1943, Abadie several genera (Selosse and Roy 2009). et al. 2006) and C. rubra (Burgeff 1954, J. P. Amardeilh One such genus is Cephalanthera Rich. that comprises et al. unpubl.). a total of ca 15 species in Europe, the Mediterranean, Asia In a few species of Cephalanthera, mycoheterotrophy and the western part of the USA. For C. damasonium has become fixed, and the permanent loss of chlorophyll (Mill.) Druce, C. longifolia (L.) Fritsch and C. rubra (L.) has been accompanied by strong leaf reduction. The best known example is the sole North American representative of the genus, C. austiniae (A. Gray) A. Heller; The review and decision to publish this paper has been taken by the above mentioned for good descriptions and illustrations, see MacDougal SE. The decision by the handling SE is shared by a second SE and the EiC. (1899, sub nom. C. oregana Rchb.f.), Luer (1975) and

460 Coleman (1995). It is less widely known that five myco- already under review. A dichotomous key to the accepted heterotrophic species of Cephalanthera have been proposed species was prepared together with a taxonomic synopsis. from Myanmar (C. pusilla (Hook. f.) Seidenf.), Laos (C. ericiflora Szlach. & Mytnik; C. exigua Seidenf.) and Yunnan (C. calcarata S. C. Chen & K. Y. Lang; C. gracilis Phylogenetic analysis S. C. Chen & G. H. Zhu). Until recently, all of these species were only known from the type specimens. There- A phylogenetic analysis was performed to assess the overall fore, it has been difficult to judge if they should be phylogenetic position of the putative Cephalanthera from recognized as five distinct species, or whether improved Thailand. A flower was collected from the population on Doi Pee Pan Nam and kept in silica gel (Chase and knowledge of their range of variation would indicate that TM two or more of them are conspecific. Hills 1991) until DNA was extracted using DNeasy On 10 Apr 2001, staff members from Queen Sirikit Plant Mini Kit, according to the manufacturer’s instruc- Botanic Garden in Chiang Mai climbed Doi Pee Pan tions. Using primers ITS1P and ITS4, the nuclear rDNA Nam, Ban Khun Lao at the Chiang Mai/Chiang Rai internal transcribed spacer (ITS) was amplified and province border in northern Thailand. In hill evergreen sequenced as in Roy et al. (2009), where also the primer forest at ca 1700 m a.s.l., dominated by trees of the sequences can be found. The sequence has been deposited families Fagaceae, Lauraceae and Theaceae, they found a at GenBank (FJ454868, Table 1). Downloaded ITS hitherto unknown population of an unidentified mycohe- sequences for Cephalanthera damasonium, C. longifolia, terotrophic Cephalanthera Á a new generic record for C. rubra and representative species belonging to other Thailand. A few years later, rumours were circulating that genera of the Neottieae were also included in the analysis, a population of similar plants had been found on Doi whereas Tropidia polystachya Sw. and Nervilia shirensis Inthanon, province of Chiang Mai. On 17 Feb 2008, Schltr. were used as outgroups (for GenBank accession Santi Watthana visited this locality together with the no., see Table 1). Sequences alignment was performed Assistant Head of Doi Inthanon National Park. They under BioEdit v. 7.3.0 (Hall 1999) and corrected by eye. retrieved the population in hill evergreen forest at ca The phylogeny was computed by maximum likelihood 1600 m a.s.l., and they confirmed that it belonged to the using PhyML v. 2.4.4 (Guindon and Gascuel 2003). same unidentified species. For this analysis, a general time-reversible model of The recent discoveries in Thailand made it possible to DNA substitution was used (GTR; Lanave et al. 1984, prepare a detailed description of the species concerned, Rodriguez et al. 1990), involving unequal base frequencies and consequently to re-evaluate species delimitation in and six types of substitution. This model of DNA sub- the genus. Furthermore, we got the opportunity to collect stitution was chosen using a series of hierarchical like- a flower for DNA-based phylogenetic analysis and to lihood-ratio tests in Modeltest 3.7 (Posada and Crandall gather the first known data on habitat requirements 1998). Base frequencies were estimated, and 10 000 and breeding system for an Asian mycoheterotrophic bootstrap replicates were performed. The phylogenetic species of Cephalanthera. tree was visualized using Treeview 1.6.6 (Page 1996).

Ecology Material and methods The habitats of the newly discovered Cephalanthera Taxonomy and distribution populations on Doi Pee Pan Nam and Doi Inthanon were studied in 2005 and 2008, respectively. The forest A detailed morphological description of the mycohe- accommodating the population on Doi Pee Pan Nam was terotrophic Cephalanthera from Thailand was prepared characterized by identifying all the trees, and at both sites a through field examination of 36 individuals on Doi Pee Pan Nam on 3 Apr 2005, and in FebÁMar 2008 it was Table 1. List of species included in the phylogenetic analysis, with ascertained that the description also covered the range of the GenBank accession numbers indicated. variation in the population on Doi Inthanon. Further- Taxon GenBank more, plants were photographed in situ, and a few accession no. individuals were collected for herbarium documentation and for preparation of a detailed line drawing. To identify Cephalanthera damasonium (Mill.) Druce AY146446 the species and to evaluate species delimitation in the Cephalanthera exigua Seidenf. FJ454868 complex in general, our description and illustrations were Cephalanthera longifolia (L.) Fritsch DQ182464 Cephalanthera rubra (L.) Rich. AY146445 then compared with descriptions (and illustrations, when- Epipactis flava Seidenf. FJ454869 ever possible) in the protologues of C. pusilla (Hooker Epipactis helleborine (L.) Crantz s.s. AY351375 1890Á1894), C. exigua (Seidenfaden 1975), C. calcarata Epipactis leptochila (Godfery) Godfery FJ454870 (Chen and Lang 1986), C. gracilis (Chen and Zhu 2002) Epipactis muelleri Godfery FJ454871 Epipactis palustris (L.) Crantz AY146448 and C. ericiflora (Szlachetko et al. 2008) and with all Limodorum abortivum (L.) Sw. AY351378 available herbarium specimens of these taxa (listed in the Neottia ovata (L.) Bluff & Fingerh. BankIt1179998 ‘Taxonomic synopsis’). Unfortunately, we did not succeed Neottia smallii (Wiegand) Szlach. AF521058 in getting the holotype of C. calcarata on loan from PE, Neottia nidus-avis (L.) Rich. AY351383 Nervilia shirensis Schltr. AF521066 and there was no time to access the type of C. ericiflora as Tropidia polystachya Sw. EU490674 this species was not described until our manuscript was

461 rectangular plot (1050 m) was demarcated for more Examination of the types of C. gracilis and C. pusilla detailed studies. and of a specimen previously identified as C. calcarata by Within each plot, samples from all vascular plant P. Ormerod (Ching 22569 K), together with information species were collected, and herbarium specimens were from the protologues of these taxa as well as C. ericiflora, prepared for later identification. Based on the identifica- did not reveal a correspondingly close match with our Thai tions, Sørensen’s (1948) presence community coefficient material. Thus, C. pusilla had a completely different was calculated as: ISS [c/½(AB)]100%, where c is labellum morphology; C. ericiflora had sepals and petals the number of species common to both plots, whereas A with subcaudate reflexed apices, different ornaments on the and B are the total numbers of species recorded in the labellum and an apiculate epichile; C. calcarata had a more plots on Doi Pee Pan Nam and Doi Inthanon, respec- distinctly set-off spur and a differently shaped epichile; and tively. In comparison to Jaccard’s original index of C. gracilis had longer inflorescences, smaller flowers and a similarity (ISJ [c/(ABc)]100%), Sørensen’s index hypochile that was saccate in its middle to distal (rather expresses the actually measured coinciding species occur- than basal) part. rences against the theoretically possible ones. As noted by Comparing C. pusilla, C. calcarata, C. gracilis and Mueller-Dombois and Ellenberg (1974), this may be C. ericiflora to each other, we found C. pusilla to have a mathematically more satisfactory, as it includes a statistical labellum morphology that was strikingly different from that probability term. in the other three taxa. Judging from the only two In both plots, all individuals of Cephalanthera were specimens available to us, C. calcarata was clearly distinct tagged during flowering in 2008 (17 Feb, 9 and 15 Mar on from C. gracilis due to its basal and more distinctly set off Doi Inthanon; 19 Mar and 9 Apr on Doi Pee Pan Nam). spur and due to its differently shaped epichile. Finally, the The number of flowers of each individual was counted, and original description and illustration of C. ericiflora suggest flowers were checked for presence of nectar. When flower- that this species is distinguished from the others in ing had ceased, the number of fruits developed on each characters pertaining to the sepals, petals and labellum (cf. individual was scored (4 Jun on Doi Inthanon; 27 Jun on key below). Doi Pee Pan Nam) in order to assess the natural levels of fruit set in the two populations. On Doi Inthanon, 17 individuals of Cephalanthera Phylogenetic analysis were covered with nylon net while still in bud. Upon anthesis, the altogether 38 flowers were experimentally An ITS sequence was successfully obtained from C. exigua, treated to test for fruit set resulting from spontaneous and the phylogenetic analysis strongly supported its position within the genus Cephalanthera (Fig. 1), thus self-pollination (n19), induced self-pollination (n11) confirming the generic identification based on morpholo- and induced cross-pollination (n8). Following the gical features. As could be expected, C. exigua is nested in experimental treatments, the nets were replaced immedi- the clade of mixotrophic Cephalanthera. ately, and the fruit set was assessed on 4 Jun.

Ecology Results The composition of the vascular plant flora in the plots Taxonomy and distribution on Doi Pee Pan Nam and Doi Inthanon, respectively, is shown in Table 2 (with additional tree species found Careful examination of the type specimen of C. exigua, outside the plot on Doi Pee Pan Nam listed in a together with information from the protologue, revealed a footnote). Because much of the collected material was in very close similarity to our Thai plants. The only differences a vegetative state, not all specimens could be identified to were a few measurements (of the bracts, labellum and species (or even genus), but in all cases it was possible to column) lying either on or slightly below the lower limits in establish the number of species from each family. The our description. Furthermore, judging from Seidenfaden’s flora was slightly more diverse in the plot on Doi (1975) description and line drawing, the epichile of Inthanon (0.086 species per m2) than in the plot on Doi C. exigua differed in being boat-shaped, rounded and Pee Pan Nam (0.056 species per m2). Furthermore, the devoid of ornaments. The smaller column length indicated floristic composition was markedly dissimilar between by Seidenfaden (1975) for C. exigua could be explained by the two plots (ISS 22.5%). the circumstance that Seidenfaden, unlike us, probably did In the plot on Doi Inthanon, a total of 126 Cepha- not include the anther in this measurement. The rather lanthera individuals (all with solitary flowering shoots) small dimensions of bracts and labellum in C. exigua were produced from 1Á3 flowers per inflorescence. The alto- probably due to moderate shrinkage of the type during the gether 183 flowers developed 8 fruits (0Á2 per inflores- drying process. Finally, we decided that the difference in cence), corresponding to a natural relative fruit set of epichile morphology between our Thai material and the 4.4%. In the plot on Doi Pee Pan Nam, a total of type of C. exigua (as interpreted by Seidenfaden) was 133 individuals produced from 1Á4 flowers per inflores- probably an artefact. This view was based on the very poor cence. The altogether 281 flowers developed 31 fruits condition of the epichile on the only flower of the type (0Á2 per inflorescence), corresponding to a natural relative specimen; we could not ourselves establish the epichile fruit set of 7.7%. No nectar production was detected in morphology of the type in detail. the flowers.

462 Figure 1. Phylogenetic tree showing the position of Cephalanthera exigua within the Neottieae. Phylogeny based on ITS using the maximum likelihood method (GTR model). Numbers on branches indicate bootstrap values above 70% (over 10 000 replicates); the scale bar shows 0.05 substitutions per site.

Concerning the bagging experiments, the levels of was apparently unaware of the type specimen at CAL) is an relative fruit set were 0% resulting from spontaneous self- accurate copy of these sketches. One of the floral sketches pollination, 72.7% resulting from induced self-pollination shows a labellum much unlike any known species of and 87.5% resulting from induced cross-pollination. Cephalanthera, and it was probably this sketch that convinced Tang and Wang (1951) that Galeola pusilla should rather be referred to Aphyllorchis Á only the basal Discussion auricles typical of an Aphyllorchis labellum are missing on the sketch. Indeed, if it could be demonstrated that the Species delimitation and systematic affinities absence of these structures on the sketch were due to the very base of the labellum being left on the column, the idea Based on our comparisons with descriptions, illustrations, of placing this species in Aphyllorchis would appear sensible. and available herbarium specimens of Asian mycohetero- However, during a visit to CAL in 2008, H. Æ. Pedersen trophic orchids previously assigned to Cephalanthera,we got an opportunity to boil and examine all parts of the confidently identify our Cephalanthera material from the Thai mountains Doi Inthanon and Doi Pee Pan Nam as dissected flower from the type specimen. This examination C. exigua (Fig. 2Á3). This represents a new generic revealed that the labellum had been sketched in its entirety and specific record for Thailand, and it considerably by Hooker, but it also revealed that Hooker’s sketch is widens the known geographic range of this species, which inaccurate. In reality, the labellum morphology suggests was previously considered endemic to Laos. Furthermore, a Cephalanthera rather than an Aphyllorchis, and this though considerably less well-substantiated, we recognize indication was supported by the completely sessile ovary the Yunnanese C. calcarata and C. gracilis and the Laotian of all three flowers (all known species of Aphyllorchis have C. ericiflora as three distinct species. a shortly pedicelled ovary). Still, the shallowly (rather When first reporting his find of the present C. exigua than deeply) three-lobed condition of the labellum in from Laos, Kerr (1933) identified his plant as Galeola Cephalanthera pusilla at once distinguishes this species pusilla Hook. f., the basionym of Cephalanthera pusilla. from the other mycoheterotrophic species of Cephalanthera When Tang and Wang (1951) transferred the latter species in Asia, including C. exigua. to Aphyllorchis Blume, they compared Kerr’s plant with the In the phylogenetic analysis that included C. exigua, protologue (Hooker 1890Á1894) and Hooker’s unpub- C. damasonium, C. longifolia, C. rubra and representative lished sketches, and they felt that Kerr was right in species of Epipactis Zinn, Limodorum Boehm. and Neottia considering the two plants as conspecific. A herbarium Guett. (Fig. 1), Cephalanthera exigua is sister to the species specimen of C. pusilla (Kurz 336) is present at CAL and pair C. damasonium/C. longifolia, and C. rubra is sister to must be regarded as the type. The sheet carries an originally the clade consisting of these three species. Apart from the three-flowered individual. Two flowers are extant on the 87% bootstrap support of the C. damasonium/C. longifolia specimen, whereas the third has been detached and clade, however, the internal relationships in Cephalanthera dissected, and the various parts of this flower are now are poorly supported. Still, the Cephalanthera clade as kept in a capsule on the sheet. The same sheet also carries a whole is well supported (100%). This strengthens Hooker’s original sketches of the plant, including floral the perception of Cephalanthera as being monophyletic details. A drawing at K (cited by Seidenfaden 1975, who (Pridgeon et al. 2005), and it corroborates the systematic

463 Table 2. List of vascular plant taxa occurring in the study plots on Doi Pee Pan Nam (A) and Doi Inthanon (B), northern Thailand. Specimens found (), not found ().

Family Species A B

Trees* Aceraceae Acer laurinum Hassk. Caprifoliaceae Viburnum sp. Clusiaceae Calophyllum sp. Fabaceae Archidendron lucidum I. C. Nielsen Fagaceae Castanopsis acuminatissima (Blume) A. DC. Fagaceae Castanopsis sp. Fagaceae Lithocarpus sp. 1 Fagaceae Lithocarpus sp. 2 Fagaceae Quercus aliena Blume Juglandaceae Engelhardtia spicata Blume Lauraceae Cinnamomum sp. Lauraceae Litsea sp. Lauraceae Neolitsea sp. Lauraceae Gen. et sp. indet. 1 Lauraceae Gen. et sp. indet. 2 Lauraceae Gen. et sp. indet. 3 Magnoliaceae Magnolia floribunda (Finet & Gagnep.) Figlar Meliaceae Heynea trijuga Roxb. ex Sims Moraceae Ficus parietalis Blume Myricaceae Myrica esculenta Buch.-Ham. Myrtaceae Syzygium angkae (Craib) P. Chantaranothai & J. Parn. Myrtaceae Syzygium sp. Theaceae Camellia connata (Craib) Craib Theaceae Camellia oleifera Abel. Theaceae Gordonia dalglieshiana Craib Theaceae Schima wallichii (DC.) Korth. Ulmaceae Gironniera sp. Xanthophyllaceae Xanthophyllum sp. Shrubs Acanthaceae Strobilanthes sp. Melastomataceae Melastoma sp. Myrsinaceae Ardisia sp. 1 Myrsinaceae Ardisia sp. 2 Myrsinaceae Maesa sp. Rubiaceae Ixora sp. Rubiaceae Lasianthus sp. Rubiaceae Psychotria sp. 1 Rubiaceae Psychotria sp. 2 Rutaceae Melicope pteleifolia (Champ. ex Benth.) Hartley Climbers Apocynaceae Gen. et sp. indet. Combretaceae Combretum sp. Myrsinaceae Embelia pulchella Mez Myrsinaceae Embelia sessiliflora Kurz Piperaceae Piper sp. Rosaceae Rubus sp. 1 Rosaceae Rubus sp. 2 Smilacaceae Smilax lanceifolia Roxb. Smilacaceae Smilax ovalifolia Roxb. Smilacaceae Smilax rigida Wall. ex Kunth Herbs Balsaminaceae Impatiens sp. Convallariaceae Aspidistra sp. Convallariaceae Ophiopogon sp. Cyperaceae Carex baccans Nees Cyperaceae Scleria sp. Dennstaedtiaceae Pteridium aquilinum Kuhn Gentianaceae Gentiana hesseliana Hosseus Lamiaceae Pogostemon sp. Phormiaceae Dianella ensifolia (L.) DC. Poaceae Bambusa sp. Poaceae Thysanolaena maxima Kuntze Pteridaceae Pteris bella Tagawa

464 Table 2 (continued)

Family Species A B

Violaceae Viola curvistylis Boissieu & Gagnep. Zingiberaceae Amomum sp. 1 Zingiberaceae Amomum sp. 2

*Additional tree taxa registered outside the plot on Doi Pee Pan Nam. Á Araliaceae: Schefflera sp.; Asteraceae: Vernonia sp.; Daphniphyllaceae: Daphniphyllum sp.; Elaeocarpaceae: Elaeocarpus sp.; Ericaceae: Rhododendron arboreum Sm., R. moulmeinense Hook.f.; Euphorbiaceae: Glochidion sp.; Fabaceae: Archidendron clyperaria (Jack) I. C. Nielsen; Fagaceae: Trigonobalanus doichangensis (A. Camus) Forman; Proteaceae: Helicia sp.; Rutaceae: Euodia triphylla DC.; Symplocaceae: Symplocos macrophylla Wall. ex DC. subsp. sulcata (Kurz) Noot.; Theaceae: Anneslea fragrans Wall., Eurya nitida Korth., Gordonia dalglieshiana Craib. position of C. exigua at the generic level. It is not yet clear Thelephoraceae should be specifically protected, and the whether mycoheterotrophy arose more than once in the habitat should be generally protected against fires and soil genus; in the slightly more comprensive phylogeny of disturbance. Pridgeon et al. (2005), the mycoheterotrophic C. austiniae assumed a position similar to C. exigua in our analysis. In a larger phylogeny that includes representatives of Aphyl- Breeding system and fruit set in Cephalanthera exigua lorchis and Cephalanthera (Roy et al. unpubl.), the former is clearly distinct, and the monophyly of Cephalanthera is With its white, zygomorphic, more or less horizontal strongly supported. flowers, its shortly spurred hypochile and its reasonably flat epichile (offering a landing platform), C. exigua fits the bee pollination syndrome of van der Pijl and Dodson Occurrence, habitat requirements and conservation (1966). Furthermore, the brown markings on the nectarless of Cephalanthera exigua labellum may be interpreted as deceptive nectar guides, whereas the ruminate ridge of the hypochile and/or the Judging from label information on Kerr’s type from Laos papillae of the epichile may represent pseudopollen. and from our own field studies in Thailand, C. exigua Only the apparently scentless condition of the flowers generally grows in hill evergreen forest from 1500Á1700 m does not fit the bee pollination syndrome (although the a.s.l. It seems to require places with deep leaf litter, but may fact that the flowers are scentless to humans does not occur both under heavy shade and in only partly shaded necessarily imply that they are scentless to insects as spots. The compositions of the vascular plant floras in well). Deceit pollination by solitary bees is well-known our plots on Doi Pee Pan Nam and Doi Inthanon are from other species of Cephalanthera. Thus, Dafni and given in Table 2. As the floras are markedly dissimilar Ivri (1981) and Vo¨th (1999) observed C. longifolia to (ISS 22.5%), C. exigua does not seem to have particularly strict habitat requirements. be pollinated by three species of Andrena (Andrenidae), Based on data collected from Doi Pee Pan Nam, Roy Halictus sp. and Lasioglossum laevigatum (Halictidae), et al. (unpubl.) found C. exigua to form mycorrhiza with whereas Nilsson (1983), Nazarov and Ivanov (1990) and basiodiomycetes of the family Thelephoraceae. This family Vo¨th (1999) observed C. rubra to be pollinated by three forms other mycorrhizal associations (i.e. ectomycorrhizae) species of Chelostoma (Megachilidae) and Dufourea denti- with forest trees that provide them their organic carbon ventris (Halictidae). In contrast, C. damasonium (with (Selosse and Roy 2009). Like all other mycoheterotrophic almost closed flowers even at the peak of flowering) is Neottieae studied so far (Taylor et al. 2002, Abadie et al. obligately autogamous (Scacchi et al. 1991, Claessens and 2006), C. exigua thus relies indirectly on trees as a carbon Kleynen 1995, Vo¨th 1999), and van der Cingel (2001) source. Species of the Thelephoraceae are regarded as non- interpreted the mycoheterotrophic C. austiniae as faculta- specific ectomycorrhizal partners of forest trees (Smith tively autogamous. and Read 2008). At both sites of C. exigua studied in The complete lack of fruit set in the 19 bagged, experi- Thailand, the vascular plant flora turned out to include mentally untouched flowers demonstrates that C. exigua several ectomycorrhizal tree species of the families Fagaceae is not spontaneously autogamous. On the other hand, and Ulmaceae (Table 2). Among the members of these the high and fairly equal levels of fruit set resulting families, only an unidentified species of Gironniera (Ulma- from induced self-pollination (72.7%) and induced cross- ceae) was found (scattered) at both study sites. At the site on pollination (83.3%) clearly indicate that this species is Doi Pee Pan Nam, Castanopsis acuminatissima (Blume) A. genetically self-compatible. The very low levels of natural DC. (Fagaceae) was found to be particularly common, and fruit set (4.4% on Doi Inthanon, 7.7% on Doi Pee Pan several species of Castanopsis (D. Don) Spach have pre- Nam) support that the nectarless flowers of C. exigua viously been reported to form ectomycorrhizae with repre- are pollinated by insects, and the figures fit the general sentatives of the Thelephoraceae (Tam and Griffiths 1993). level of 0.0Á43.0% relative fruit set in nectarless tropical The indirect nutritional dependency of C. exigua on orchids (compared to 17.8Á41.0% in nectar-producing ectomycorrhizal trees, as represented by members of the species of this group), confer Neiland and Wilcock Fagaceae and Ulmaceae at the two sites in Thailand, should (1998). Incidentally, the natural level of mean relative be taken into consideration when setting up conservation fruit set has been reported to vary from 17.3Á62.4% in measures for this very rare species. Thus, at each locality C. longifolia (Dafni and Ivri 1981) and from 0.0Á30.1% of C. exigua, the trees forming ectomycorrhizae with in C. rubra (Nilsson 1983).

465 Figure 2. Cephalanthera exigua Seidenf. (A) habit, (B) bract, (C) ﬂower, (D) ﬂower (sepals and petals removed), (E) dorsal sepal, (F) lateral sepal, (G) petal, (H) labellum, (J) column. Double-line scale1 cm. Single-line scale1 mm. (A)Á(J) Watthana 1338 C. Henrik Æ. Pedersen del.

Based on the lack of nectar, the patterns of fruit set Á Labellum deeply three-lobed near the middle, dis- and the pollination syndrome (supported by previous tinctly differentiated in hypochile and epichile . . . . 2 anthecological studies of other Cephalanthera species), we 2. Sepals and petals with subcaudate, reflexed apices. conclude that C. exigua has an insect-operated breeding Epichile broadly rounded, strongly apiculate ...... system acting by deceit, and that the pollinators are ...... 4.C. ericiflora probably small solitary bees. Á Sepals and petals with obtuse to acuminate, straight to slightly recurved apices. Epichile rounded to acute, Taxonomic synopsis not apiculate ...... 3 3. Hypochile with a conical, obtuse, distinctly set off Key to the Asian mycoheterotrophic species of spur at base; epichile transversely elliptic, rounded . . Cephalanthera ...... 3.C. calcarata Á Hypochile gradually widened into a rounded sac; 1. Labellum shallowly three-lobed in its distal part, not epichile ovate to cordate, obtuse ...... 4 distinctly differentiated in hypochile and epichile . . . 4. Inflorescence up to 4 cm long (measured from the node ...... 5.C. pusilla of the lowermost flower). Sepals at least 11 mm long.

466 shaped band near the apex of the epichile; immovably attached to the column, porrect, differentiated in hypochile and epichile; hypochile saccate with erect, falcately triangular, obtuse to subacute sidelobes and a median ruminate ridge, glabrous and smooth (except for the slightly papillose apices of the sidelobes), 2.9Á4.6 mm long, 9.5Á14.0 mm across the sidelobes when flattened; sac 2.0Á3.5 mm long (measured along the back wall); epichile cordate with upcurved sides and downcurved apex, obtuse, densely papillose and carrying 3Á5 longitudinal keels on the ventral side, 4.3Á8.34.9Á9.9 mm. Column semiterete, straight to slightly incurved, 8.0Á9.6 mm long including the anther; anther versatile, obliquely ellipsoidal, obtuse to subacute, 2-loculate with white, strongly elongate, mealy pollinia, 2.7Á2.8 mm long; rostellum well-developed, protruding below the anther; receptive part of stigma concave with protruding margins. Ovary terete, striated, 11.6Á12.6 mm long.

Additional specimen examined Thailand, northern floristic region, Chiang Mai, Doi Inthanon, 1600 m a.s.l., 17 Feb 2008 (Watthana 2714 QBG!), Doi Pee Pan Nam, 1700 m a.s.l., 10 Apr 2001 (Watthana 1338 BKF! C! QBG!), 3 Apr 2005 (Suddee et al. Figure 3. Cephalanthera exigua Seidenf. Thailand, northern ﬂoristic region, Chiang Mai, Doi Pee Pan Nam, 1700 m a.s.l., 2181 BKF!). 23 Mar 2005, photo: Henrik Æ. Pedersen. Hypochile saccate at base. Column (including anther) Cephalanthera gracilis S. C. Chen & G. H. Zhu (2002, at least 8 mm long ...... 1. C. exigua p. 600) Á Inflorescence more than 4 cm long (measured from the node of the lowermost flower). Sepals up to 10 mm Type: China, Yunnan, Binchuan county, 1 May 1911, long. Hypochile saccate in its middle part. Column Ducloux 7171 (holotype: P!). (including anther) up to 6 mm long . . . 2. C. gracilis Cephalanthera calcarata S. C. Chen & K. Y. Lang (1986, p. 271) Cephalanthera exigua Seidenf. (1975, p. 71) (Fig. 2, 3) Type: China, Yunnan, Yangbi county, Ji Dan Shan west of Type: Laos, Xiangkhoang, ‘Phu Muten’, 1500 m a.s.l., 20 Cang Shan, near Miao Ju Ping, 7 May 1929, Ching 22569 Apr 1932, Kerr 01024 (holotype: K!). (holotype: PE, isotype: KUN). Plant white (sometimes with a faint greenish tinge in backlight), rhizomatous with a few, unbranched, slenderly Additional specimen examined fusiform roots; flowering shoots 3.0Á14.0 cm tall. Stem China. Yunnan, Yangbi county, above Yangbi, 2600 m slender, 1Á2 mm in diameter; internodes 1.2Á4.4 cm. a.s.l., 9 May 1981 (Sino-British Expedition to Cangshan Leaves 1Á2 (excluding 4Á6 underground cataphylls), 1981 0397 K!). each of them reduced to a 0.6Á1.0 cm long sheath and a boat-shaped, (ovateÁ)triangular, acute to obtuse lamina Cephalanthera ericiflora Szlach. & Mytnik (in Szlach. (0.3Á0.90.2Á0.6 cm). Inflorescence 0.9Á4.0 cm long, et al. 2008, p. 213) 1- to 4-flowered (rarely 5-flowered), often somewhat Á secund. Floral bracts narrowly lanceolate oblong, acute, Type: Laos, Viangchan, ‘Pu Tat’, 1200 m a.s.l., 22 Apr Á Á 12.0 13.9 3.0 4.2 mm, 5- to 7-veined from the base. 1932, Kerr s.n. (holotype: BM). Flowers sessile, resupinate, suberect (to spreading). Sepals white, somewhat spreading, glabrous and smooth; dorsal sepal lanceolate, subacuminate, 3- to 5-veined from the Cephalanthera pusilla (Hook. f.) Seidenf. (1975, base, 11.0Á21.03.9Á5.7 mm; lateral sepals lanceolate, p. 72) slightly oblique, somewhat channelled towards the apex, acuminate, 3- to 4-veined from the base, 12.0Á20.1 Basionym: Galeola pusilla Hook. f. (1890Á1894, p. 89 4.1Á6.1 mm. Petals white, porrect, glabrous and smooth, [published 1890]). narrowly ellipticÁoblong, slightly recurved at the apex, Based on the same type: Aphyllorchis pusilla (Hook. f.) obtuse, 3-veined from the base, 9.3Á15.53.1Á5.2 mm. Tang & F. T. Wang (1951, p. 66) p.p. Labellum white with pale brown ornaments, pale brown Type: Myanmar, Bago, Pegu, ‘on the Pookee ridges’, sine inner surface of the hypochile and a pale brown crescent- anno, Kurz 336 (holotype: CAL!; drawing at K!).

467 Acknowledgements Á H. Æ. Pedersen is indebted to the Carlsberg Molvray, M. et al. 2000. Polyphyly of mycoheterotrophic orchids Foundation for financial support. Additionally, we wish to thank and functional influences on floral and molecular characters. the curators of CAL, K and P for their help and hospitality, Mikael Á In: Wilson, K. L. and Morrison, D. A. (eds), Monocots: Hedreń for useful comments on the manuscript and the Socie´te´ systematics and evolution. CSIRO, pp. 441Á448. Française d’Orchidophilie (SFO) for support to M.-A. Selosse and Mueller-Dombois, D. and Ellenberg, H. 1974. Aims and methods M. Roy. of vegetation ecology. Á John Wiley. Nazarov, V. V. and Ivanov, S. P. 1990. Ucˇastie pcˇel roda Chelostoma Latr. (Hymenoptera, Megachilidae) v opylenii References mimikrirujusˇcˇix vidov Cephalanthera rubra (Z.) Rich. i Campanula taurica Juz. v Krymu. Á Eńtomologicheskoe Abadie, J.-C. et al. 2006. Cephalanthera longifolia (Neottieae, Obozrenie LXIX: 534Á537. Orchidaceae) is mixotrophic: a comparative study between Neiland, M. R. M. and Wilcock, C. C. 1998. Fruit set, nectar green and nonphotosynthetic individuals. Á Can. J. Bot. 84: reward, and rarity in the Orchidaceae. Á Am. J. Bot. 85: 1657Á 1462Á1477. 1671. Burgeff, H. 1954. Samenkeimung und Kultur europaïscher Nilsson, L. A. 1983. Mimesis of bellflower (Campanula) by the Erdorchideen nebst Versuchen zu ihrer Verbreitung. Á Gustav red helleborine orchid Cephalanthera rubra. Á Nature 305: Fischer. 799Á800. Chase, M. W. and Hills, H. G. 1991. Silica gel: an ideal material Page, R. D. M. 1996. Treeview: an application to display for field preservation of leaf samples for DNA studies. Á Taxon phylogenetic trees on personal computers. Á Computer Appl. 40: 215Á220. Biosci. 12: 357Á358. Chen, S.-C. and Lang, K.-Y. 1986. Cephalanthera calcarata, a new van der Pijl, L. and Dodson, C. H. 1966. Orchid flowers: their saprophytic orchid from China [in Chinese]. Á Acta Bot. pollination and evolution. Á Fairchild Trop. Gard., Univ. of Yunnan. 8: 271Á274. Miami Press. Chen, S.-C. and Zhu, G.-H. 2002. Cephalanthera gracilis Posada, D. and Crandall, K. A. 1998. Modeltest: testing the model (Orchidaceae), a new species from China. Á Acta Bot. Yunnan. of DNA substitution. Á Bioinformatics 14: 817Á818. 24: 600Á602. Pridgeon, A. M. et al. (eds) 2005. Genera Orchidacearum 4. van der Cingel, N. A. 2001. An atlas of orchid pollination. Epidendroideae, part one. Á Oxford Univ. Press. America, Africa, Asia and Australia. Á A. A. Balkema. Rasmussen, H. N. 1995. Terrestrial orchids from seed to Claessens, J. and Kleynen, J. 1995. Die Systematik der euro- mycotrophic plant. Á Cambridge Univ. Press. Á païschen Orchideen illustriert an Hand von Makro-Fotos. J. Renner, O. 1938. U¨ ber blasse, saprophytische Cephalanthera alba Eur. Orchid. 27: 93Á124. und Epipactis latifolia. Á Flora 132: 225Á233. Coleman, R. A. 1995. The wild orchids of California. Á Cornell Renner, O. 1943. Notiz u¨ber blasse Cephalanthera und Tozzia. Univ. Press. Á Flora 136: 309Á312. Dafni, A. and Ivri, Y. 1981. The flower biology of Cephalanthera Rodriguez, F. et al. 1990. The general stochastic model of longifolia (Orchidaceae) Á pollen imitation and facultative floral mimicry. Á Plant Syst. Evol. 137: 229Á240. nucleotide substitution. Á J. Theor. Biol. 142: 485Á501. Gebauer, G. and Meyer, M. 2003. 15N and 13C natural Roy, M. et al. 2009. Ectomycorrhizal Inocybe species associate with abundance of autotrophic and mycoheterotrophic orchids the mycoheterotrophic orchid Epipogium aphyllum but not provides insight into nitrogen and carbon gain from fungal with its asexual propagules. Á Ann. Bot. Oxford 104: 595Á610. association. Á New Phytol. 160: 209Á223. Scacchi, R. et al. 1991. Effect of the breeding system on the Guindon, S. and Gascuel, O. 2003. A simple, fast, and accurate genetic structure in three Cephalanthera spp. (Orchidaceae). algorithm to estimate large phylogenies by maximum like- Á Plant Syst. Evol. 176: 53Á62. lihood. Á Syst. Biol. 52: 696Á704. Seidenfaden, G. 1975. Contributions to the orchid flora of Hall, T. A. 1999. BioEdit: a user-friendly biological sequence Thailand VI. Á Bot. Tidsskr. 70: 64Á97. alignment editor and analysis program for Windows 95/98/ Selosse, M.-A. and Roy, M. 2009. Green plants that feed on fungi: NT. Á Nucl. Acids Symp. Ser. 41: 95Á98. facts and questions about mixotrophy. Á Trends Plant Sci. 14: Hooker, J. D. 1890Á1894. The flora of British India VI. Orchideæ 64Á70. to Cyperaceæ. Á L. Reeve and Co. Smith, S. E. and Read, D. J. 2008. Mycorrhizal symbiosis (3rd Julou, T. et al. 2005. Mixotrophy in orchids: insights from a ed.). Á Academic Press. comparative study of green individuals and nonphotosynthetic Sørensen, T. 1948. A method of establishing groups of equal individuals of Cephalanthera damasonium. Á New Phytol. 166: amplitude in plant sociology based on similarity of species 639Á653. content. Á Biol. Skr. 5: 1Á34, Fig. 1, Table 4 (1Á6). Kerr, A. F. G. 1933. A collection of orchids from Laos. Á J. Siam Szlachetko, D. L. et al. 2008. Cephalanthera ericiflora Szlach. & Soc., Nat. Hist. Suppl. IX: 225Á243. Mytnik, sp. nov. (Orchidaceae, Neottioideae), a new species Kohns, P. and Schneider, P. 1993. Cephalanthera damasonium from Laos. Á Acta Soc. Bot. Poloniae 77: 213Á215. (Mill.) Druce var. chlorotica Á ein Standort im Saarland. Tam, P. C. F. and Griffiths, D. A. 1993. Mycorrhizal associations Á Orchidee (Hamburg) 44: 31Á32. in Hong Kong Fagaceae. Á Mycorrhiza 2: 111Á115. Lanave, C. et al. 1984. A new method for calculating evolutionary Tang, T. and Wang, F. T. 1951. Contributions to the know- substitution rates. Á J. Mol. Evol. 20: 86Á93. ledge of eastern Asiatic Orchidaceae II. Á Acta Phytotax. Sin. 1: Leake, J. R. 1994. The biology of myco-heterotrophic (‘sapro- 23Á102. phytic’) plants. Á New Phytol. 127: 171Á216. Taylor, D. L. et al. 2002. Mycorrhizal specificity and function in Luer, C. A. 1975. The native orchids of the United States and myco-heterotrophic plants. Á In: van der Heijden, M. G. A. Á Canada excluding Florida. NY Bot. Gard. and Sanders, I. R. (eds), Mycorrhizal ecology. Ecol. Stud. MacDougal, D. T. 1899. Symbiosis and saprophytism. Á Bull. Analysis Synth. 157. Springer, pp. 375Á413 (IÁXXIV, 1Á469). Torrey Bot. Club 26: 511Á530, Pls 367Á369. Vo¨th, W. 1999. Lebensgeschichte und Bestaüber der Orchideen Mairold, F. and Weber, F. 1950. Notiz u¨ber Cephalanthera am Beispiel von Niederosterreich. Á Stapfia 65: 1Á257. Albinos. Á Protoplasma 39: 275Á277. ¨

468 79.(1*= 79.(1*= = = = _= `=_8_= _= (414,.*=)*=6:*16:*8=47(-.)S*8=+47*89.T7*8=a=*8=(1*+8=54:7=1*:7= (438*7;&9.43 _=3/*:==)*=(438*7;&9.43=54:7=1*8=47(-.)S*8=(&:88*3&7)*8_=(9*8=)*=1&=/4:73S*= 8(.*39.+.6:*=):=5&7(=)*8= 7&3)8= &:88*8`=.11&:`=8*59*2'7*=,**1_= = = *9=&79.(1*=*89=.88:=):3*=(43+S7*3(*=)433S*=1478=):3*=/4:73S*=)*=7*(4397*8=47,&3.8S*=5&7=1*= 5&7(=)*8= 7&3)8=(&:88*8=A=.11&:_=)1=8&,.9=):3=&79.(1*=)*=;:1,&7.8&9.43=54:7=8*38.'.1.8*7=1*8= ,*89.433&.7*8= )*= 5&7(8= *9= 1*8= 47(-.)45-.1*8= A= 1&= 3S(S88.9*= )*= (438.)S7*7= 1*8= 7*1&9.438= 2>(47-.?.*33*8=)&38=1*8=574/*98=)*=(438*7;&9.43_= = ,&.=.3(1:8=(*9=&79.(1*=)&38=243=2S24.7*=54:7=9S24.,3*7=)*=2&=)S2&7(-*=)*=;:1,&7.8&9.43_= ,&.=5&7=&.11*:78=5:'1.S=)*:==&:97*8=&79.(1*8=)*=;:1,&7.8&9.43=a== = _= `=_8_= =,**1_= *8=2>89T7*8=)*8=47(-.)S*8=+&392*8_=2:11*9.3=)*=1&=4(.S9S= 37&3O&.8*=) 7(-.)45-.1.*=):= &3,:*)4(854:88.1143_=6S(*2'7*=,**1_= = _= 5 7= ,**2_=93*=-.894.7*= 6:.= 54:77&.9=+&.7*=7U;*7= '.*3= )*8= (-&25.,3438c *8= 2>89T7*8= )*8=47(-.)S*8=+&392*8_=<1:2*=2`=5_=+-_= 61

Ecologie de quelques orchidées forestières : des clefs pour leur conservation

Mélanie ROY* & Marc-André SELOSSE*

Centre d’Ecologie Fonctionnelle et Evolutive, équipe Interactions Biotiques. CNRS, 1919, route de Mende. 34293 Montpellier cedex 5 [email protected] * membres de la SFO

Résumé Les orchidées forment une des plus grandes familles de plantes à fleur. En France, il existe 187 espèces et toutes sont terrestres. Certaines n’ont ni feuille ni chlorophylle, comme la Néottie nid-d’oiseau (Figure 1 à gauche) ou l’Epipogon sans feuille. Elles ne produisent pas leur propre nourriture (la matière organique), et sont dites hétérotrophes. Leur nutrition est longtemps restée un mystère qui fut percé seulement ces dernières années. Nous présentons ici les exigences écologiques de quelques espèces, comme la Néottie nid-d’oiseau, mais aussi de plusieurs autres orchidées vertes de nos forêts comme la Céphalanthère de Damas (Figure 1 à droite). Nous discutons de la nécessaire prise en compte de ces exigences dans les programmes de conservation.

Figure 1. Deux orchidées de sous-bois, à gauche la Néottie nid-d’oiseau ( Neottia nidus-avis ), une orchidée terrestre sans feuille verte, et à droite la Céphalanthère de Damas, une orchidée terrestre chlorophyllienne, (Cephalanthera damasonium ). (Photos : M. Roy)

1. De l’importance des champignons dans la vie des orchidées La plupart des orchidées terrestres et de toutes les orchidées épiphytes ont des feuilles vertes développées et produisent leur propre matière organique à partir de la photosynthèse à l’âge adulte. Cependant, les orchidées produisent des graines minuscules (environ 50 µm de long) qui sont dépourvues de réserve et qui ne peuvent pas assurer les premières étapes de la germination. Elles ne peuvent germer qu’en présence de champignons particuliers, les rhizoctonias, connus aussi comme saprophytes (= qui dégradent la matière morte du sol) ou 62 parasites sur d’autres plantes et fréquents en milieu ouvert. Ces champignons ne forment pas un groupe monophylétique (= pas un groupe homogène avec une seule origine), et sont représentés par trois genres : Ceratobasidium , Tulasnella et les Sebacina de type B.

Le champignon pénètre dans la graine et même dans les cellules de l’orchidée, où il grandit, occupe la majeure partie de la cellule et forme des « pelotons » de filaments (Figures 2). Ces champignons lèvent la dormance de la graine et assurent sa nutrition en lui fournissant des sucres, des acides aminés, de l’eau et des sels minéraux. Comme les orchidées ne produisent aucun sucre à cette étape, elles sont hétérotrophes et comme les champignons constituent leur source de carbone, ces orchidées sont dites mycohétérotrophes (= qui se nourrit de champignon).

Figure 2. A : Coupe transversale de racine d’orchidée, les champignons (en bleu) forment des pelotons dans les cellules (Photo M.A Selosse et al.) ; B : Coupe transversale de racine d' Epipactis atrorubens d’orchidée en microscopie électronique à balayage montrant les pelotons dans les cellules (Photo de T. Malonova).

Toutes les espèces de Rhizoctonia ne peuvent pas faire germer n’importe quelle espèce d’orchidée ; il existe différents niveaux de spécificité dans cette association.

A l’âge adulte, les orchidées sont toujours associées à des champignons. Il s’agit toujours de Rhizoctonia, qui forment des pelotons et qui fournissent de l’eau et des sels minéraux à l’orchidée. Mais à l’âge adulte, ces orchidées réalisent la photosynthèse et au moins dans certains cas, transfèrent une partie des sucres produits aux champignons dans leurs racines (figure 4). Ce type de relation entre l’orchidée et ses champignons est une symbiose, et fonctionne comme celle entre les chênes et les truffes par exemple, à la différence près que les truffes et les racines de chêne s’associent en une 1 mm structure différente, composée à la fois des cellules de l’arbre et du champignon, l’ectomycorhize (Figure 3). Figure 3. Ectomycorhize de Lactaire sur une racine de Pin (photo D. Mousain). 63

Énergie lumineuse

H20 H20 Sels Sels Rhizoctonia orchidée Phosphates Organique P Glutamine Minéral Acides Aminés Sucre Organique N Minéral Vitamines, Hormones

Figure 4. Relations symbiotiques existant entre la plupart des orchidées et leurs champignons associés, des rhizoctonias, à l’âge adulte. (Photo P.-H. Fabre)

2. Les mystères de la Néottie nid-d’oiseau La Néottie nid-d’oiseau ne réalise jamais la photosynthèse et ne peut donc pas donner de sucres à ses champignons associés, elle reste donc mycohétérotrophe non seulement à l’état de graine mais aussi à l’état adulte. Elle ressemble aux plantes parasites comme l’Orobanche mais n’ont pas, comme celles-ci, de racines connectées aux racines d’une autre plante : nous allons voir qu’elles se nourrissent de champignons. La forme des racines de la Néottie lui ont donné son nom et sont en effet remplies de pelotons de champignons. Du XIXème au XXème siècle, de nombreux chercheurs ont tenté de déterminer quels étaient les champignons des racines de la Néottie, car ils ne ressemblaient pas aux rhizoctonias habituels, mais la plupart des études sont restées vaines. En effet, pour identifier un champignon, la technique classique consistait à isoler le champignon des racines en prélevant les pelotons et en les cultivant sur un gel riche en sucres. Ensuite, l’observation de la croissance, de la morphologie, de la couleur de la colonie de champignon permet de l’identifier. Cependant, cette technique s’est révélée infructueuse avec la Néottie nid-d’oiseau : seuls des contaminants étaient obtenus ! Dans les années 90, des chercheurs ont trouvé une nouvelle méthode pour identifier les champignons sans observer la morphologie, en utilisant uniquement la biologie moléculaire. Il suffit d’extraire l’acide désoxyribonucléique (ADN) d’un champignon, d’amplifier un fragment de l’ADN ribosomal (ITS) et de séquencer ce fragment (Gardes & Bruns, 1993). Une fois la séquence obtenue, il est alors possible de la comparer à des banques de données et d’identifier l’espèce ou le genre de champignon. Ceci est possible car le fragment cible, l’ITS, a une séquence qui varie peu ou pas entre les individus d’une même espèce, mais varie plus entre les individus d’espèces différentes.

Cette technique a été utilisée sur la Néottie nid-d’oiseau, en extrayant l’ADN des racines. Les champignons qui ont été identifiés à partir de ces racines étaient tous des Sebacina de type A (Figure 5), différents des Sebacina de type B habituellement trouvés dans les racines d’orchidées (Selosse et al , 2002). Dans le sol autour des racines de Néottie, de nombreuses ectomycorhizes de chênes et d’autres arbres ont été trouvées : les champignons 64 identifiés dans ces ectomycorhizes étaient les mêmes Sebacina que dans les Néotties voisines ! Parmi les Sebacina , il y a deux groupes, l’un (A) qui comporte les Sebacina qui s’associent à la Néottie et aux racines des arbres, et l’autre (B) composé de saprophytes de milieu ouvert et dont certains s’associent aux racines d’orchidées de prairies (Figure 5).

Sebacina de type A identifiés chez la Néottie, ectomycorhiziens

Sebacina de type B 5 mm identifiés chez des orchidées terrestres vertes

Figure 5. Classification des Sebacina : les espèces trouvées dans la Néottie nid-d’oiseau sont plus proches de celles associées aux racines d’arbres que de celles trouvées dans les racines d’autres orchidées. A droite, une photo de carpophore (fructification) de Sebacina de type A (photo M.A. Selosse).

Le plus étonnant est que la Néottie soit associée aux mêmes champignons que les arbres, non seulement aux mêmes espèces mais aussi aux mêmes individus. Il existe donc un lien physique entre les racines de la Néottie et celles du chêne, réalisé par le mycélium du champignon. D’autre part, la Néottie reçoit sa matière organique des Sebacina qui se trouvent dans ses racines. Or, les champignons sont eux aussi incapables de produire leur propre matière organique. Les champignons symbiotiques, dont les ectomycorhiziens, utilisent la matière organique de l’organisme auquel ils sont associés. Les Sebacina associés aux arbres utilisent donc de la matière organique (ici des sucres) qui a été produite initialement par le chêne. Les Néotties associées aux Sebacina reçoivent donc la matière organique des arbres par le biais des Sebacina (Figure 6). Ce mode de nutrition est inattendu car la Néottie était autrefois considérée comme saprophyte car on pensait qu’elle se nourrissait de la matière organique du sol ; or la source de carbone qu’elle utilise provient en fait d’un arbre !

H20 H20 Sels Sels Sebacina orchidée Phosphates Phosphates

Glutamine, Glutamine, Acides Aminés Acides Aminés Carbone Sucre organique Vitamines hormones

Figure 6. Échanges et transferts d’eau de sels minéraux et de sucres entre les arbres, les Sebacina et les Néotties associées. La connexion entre les différents partenaires est permise grâce au mycélium des Sebacina .

Le transfert de matière organique entre espèces différentes par le biais de champignon était déjà connu depuis plus longtemps dans d’autres cas. L’équipe de S. Simard a montré en 65 suivant du carbone radioactif que du carbone passait d’un bouleau à un jeune sapin de Douglas par le biais de champignons ectomycorhiziens (Simard et al , 1997 ; Selosse et al , 2006). Les champignons ectomycorhiziens sont souvent généralistes et peuvent s’associer à plusieurs espèces d’arbres en même temps, de même que les arbres peuvent être associés à plusieurs espèces de champignons en même temps. Ces champignons peuvent former un réseau qui connecte les arbres d’espèces différentes, il s’agit d’un « réseau mycorhizien commun ».

La Néottie, avec son mode de nutrition particulier, est donc un puits de carbone pour ce réseau dont elle profite. Cela signifie aussi qu’elle en dépend complètement pour sa nutrition, d’autant plus qu’elle est très spécifiquement associée aux Sebacina . En termes de conservation, cela signifie que la Néottie ne peut être transplantée, ni déterrée, car cela détruirait le lien avec les arbres autour. De même, couper les arbres qui entourent la Néottie revient à lui couper les vivres et détruire la population de Néottie.

La Néottie n’est pas un cas unique, d’autres recherches ont révélé que l’espèce Corallorhiza trifida était elle aussi associée à des champignons ectomycorhiziens, des Théléphores. Il en est de même de l’espèce Cephalanthera austinae (absente en France) et même Monotropa hypophytis , une Ericacée mycohétérotrophe. De telles recherches s’avèrent donc essentielles pour comprendre le fonctionnement de nos forêts et pour la conservation des espèces mycohétérotrophes.

3. Ecologie des proches parents de la Néottie nid-d’oiseau

La Néottie nid-d’oiseau est associée à des champignons très différents de ceux associés aux autres orchidées terrestres, différents par leur écologie et par leur identité. Comment cette transition s’est elle produite ? Est-elle récente ? Pour étudier l’évolution de ce phénomène, il faut se pencher sur les racines des plus proches parents de la Néottie, comme par exemple les Epipactis ou les Céphalanthères.

L’étude de la Céphalanthère de Damas (Figure 1) a révélé qu’elle était elle aussi associée à des champignons ectomycorhiziens, mais à plusieurs espèces en même temps (Julou et al , 2005). Le cortège est donc moins spécifique, mais l’écologie est la même en terme de champignons associés. Le fonctionnement de cette association est-il le même pour autant ?

Pour le savoir, la ‘signature isotopique’ des Céphalanthères de Damas a été mesurée. Cette signature est le ratio entre les isotopes du carbone ( 13 C et 12 C) dans la plante, comparé au ratio d’une référence. Cette signature est différente entre les plantes (qui font de la photosynthèse) et les champignons par exemple. Dans certains cas, cette signature se transmet dans les chaines alimentaires : ainsi, un organisme hétérotrophe a la signature de ce qu’il mange. Par exemple, la signature de la Néottie nid-d’oiseau est la même que celle des champignons ectomycorhiziens. La signature de la Céphalanthère de Damas n’est pas exactement la même que celle des autotrophes (plantes qui font la photosynthèse), ni la même que celle que des champignons ectomycorhiziens. Elle est intermédiaire, ce qui signifie qu’elle utilise deux sources de matière organique, d’une part le carbone issu de sa propre photosynthèse et d’autre part le carbone issu des champignons ectomycorhiziens, donc celui des arbres associés. La Céphalanthère de Damas est donc dite mixotrophe, car elle utilise deux sources de matière organique à la fois (figure 7). 66

Énergie lumineuse

H 0 H20 2 Sels Sels Champignons minéra ectomyco- minéra Phosphates rhiziens Phosph

Glutamine, Glutamine, Acides Aminés Acides Aminés Sucre Carbone Vitamines, hormones

Figure 7. Échanges et transferts d’eau de sels minéraux et de sucres entre les arbres, leurs champignons ectomycorhiziens et les Céphalanthères de Damas associés. La connexion entre les différents partenaires est permise grâce aux mycéliums des champignons.

Par ailleurs, chez cette espèce, il existe une autre preuve de leur nutrition en partie mycohétérotrophe. Certains rares individus de Céphalanthères de Damas sont étonnamment dépourvus de chlorophylle. Il s’agit bien de la même espèce que les individus verts, mais avec 100 fois moins de chlorophylle et aucune photosynthèse (Figure 8). Ces individus sont dits hypochromes ou albinos, et étonnent car ils survivent, produisent des fruits, et ne diffèrent en rien (hormis la couleur) des individus verts. La signature isotopique de ces individus montre qu’ils ont la même que les champignons ectomycorhiziens. Ces individus reçoivent donc uniquement le carbone des champignons. Ceci montre l’importance du carbone issu du réseau mycorhizien, car seul le carbone des champignons suffit à former une plante, lui permettre de faire des fruits et de réapparaitre l’année suivante. D’ailleurs, sur la base de la signature isotopique des individus verts, on peut montrer que, chez la Céphalanthère de Damas verte, jusqu’à 85% du carbone provient des champignons associés.

Figure 8. A gauche un individu hypochrome de l’espèce Cephalanthera damasonium (photo M. Roy), et à droite un individu hypochrome et un autre vert de l’espèce C. longifolia (photo, U. Puttstepp), 67

Figure 9. Individus hypochromes (sans chlorophylle), à gauche, chez Epipactis helleborine (photo A. Soulié), à droite chez E. microphylla (photo G. Scappaticci).

Le déterminisme de l’hypochromie n’est pas encore connu, mais nous savons que ce phénomène est connu en région tempérée, en Amérique et en Eurasie, chez cinq espèces de Céphalanthères ( C. damasonium, C. longifolia, C. rubra, C. kurdica, C. longibracteata ) et sept Epipactis ( E. helleborine, E. purpurata, E. microphylla, E. atrorubens, E. distans, E. tremolsii, E. neerlandica ). Ces espèces sont donc toutes mixotrophes, ce qui montre que la plupart des orchidées de nos forêts sont mixotrophes, et dépendent en partie de champignons ectomycorhiziens, et donc d’arbres pour leur nutrition (Selosse et al , 2004 ; Abadie et al , 2006).

4. Conséquences pour la conservation des orchidées forestières

L’étude de la Néottie nid-d’oiseau et d’autres orchidées forestières a révélé que toutes étaient associées à des champignons ectomycorhiziens (par ailleurs associés, quant à eux, aux arbres voisins). Ces partenaires leur fournissent de la matière organique. Ce lien trophique entre les orchidées, les champignons et les arbres est essentiel pour ses orchidées et les rend d’autant plus fragiles, puisqu’elles sont dépendantes de la présence et de la santé de chacun de ces partenaires. Ceci explique aussi les essais infructueux de culture ou même de transplantation ces orchidées. Ceci révèle aussi que, pour les orchidées forestières, la présence d’arbres est essentielle, alors qu’au contraire il vaut probablement mieux éclaircir leur milieu pour les orchidées de prairies. Ces travaux ont porté principalement sur les Néottieae, mais d’autres sont en cours pour élargir cette étude à d’autres orchidées de la flore française, tel que l’Ophrys mouche, que l’on trouve souvent en lisière de forêt.

Remerciements Ces travaux ont été permis grâce à l’aide de tous les orchidophiles qui nous aident sur le terrain à recueillir des échantillons, des données, et sans qui ces travaux ne seraient pas possibles. Par ailleurs, ces recherches ont été financées par le CNRS et par la Société Française d’Orchidophilie. 68

Bibliographie

Abadie J.C., Puttsepp U., Gebauer G. , et al. 2006. Cephalanthera longifolia (Neottieae, Orchidaceae) is mixotrophic: a comparative study between green and nonphotosynthetic individuals. Canadian Journal of Botany-Revue Canadienne De Botanique 84, 1462-1477. Bournieras M., Prat D. et al (Collectif SFO), 2005. Les Orchidées de France, Belgique et Luxembourg (2 e Edition). Collection Parthénope, Biotope, Mèze, 504 p. Gardes M. & Bruns T.D. 1993. Its primers with enhanced specificity for basidiomycetes - application to the identification of mycorrhizae and rusts. Molecular Ecology 2, 113- 118. Julou T., Burghardt B., Gebauer G. , et al. 2005. Mixotrophy in orchids: insights from a comparative study of green individuals and nonphotosynthetic individuals of Cephalanthera damasonium . New Phytologist 166, 639-653. Selosse M.A., Weiss M., Jany J.L. & Tillier A. 2002. Communities and populations of sebacinoid basidiomycetes associated with the achlorophyllous orchid Neottia nidus- avis (L.) LCM Rich. and neighbouring tree ectomycorrhizae. Molecular Ecology 11, 1831-1844. Selosse M.A., Faccio A., Scappaticci G. & Bonfante P. 2004. Chlorophyllous and achlorophyllous specimens of Epipactis microphylla (Neottieae, Orchidaceae) are associated with ectomycorrhizal septomycetes, including truffles. Microbial Ecology 47, 416-426. Selosse M.A., Richard F., He X.H. & Simard S.W. 2006. Mycorrhizal networks: des liaisons dangereuses? Trends in Ecology & Evolution 21, 621-628. Simard S.W., Perry D.A., Jones M.D. , et al. 1997. Net transfer of carbon between ectomycorrhizal tree species in the field. Nature 388, 579-582. Wei ! M., Selosse M.A., Rexer K.H., Urban A. & Oberwinkler F. 2004. Sebacinales: a hitherto overlooked cosm of heterobasidiomycetes with a broad mycorrhizal potential. Mycological Research 108, 1003-1010. S8:2S= *8=51&39*8= 2>(4-S9S749745-*8=8439=343=(-14745->11.*33*8=*9=7*O4.;*39=1*:7=2&9.T7*=47,&3.6:*=)*=1*:78= (-&25.,3438=2>(47-.?.*38=*3=8>2'.48*=&;*(=1*:78=7&(.3*8_=-*?=1*8=47(-.)S*8`=1&=2>(4-S9S749745-.*=*89= &55&7:= .3)S5*3)&22*39= ,*= +4.8_= #4:9*8= 1*8= 47(-.)S*8= (-14745->11.*33*8= 8439= 51:8= 4:= 24.38= 85S(.+.6:*2*39= &884(.S*8= A= )*8= (-&25.,3438= 2>(47-.?.*38`= (433:8= (422*= 8&5745->9*8`= 9&3).8= 6:*= 1*8= *85T(*8= 2>(4-S9S749745-*8= &(9:*11*2*39= (433:*8= 8439= 85S(.+.6:*2*39= &884(.S*8= A= )*8= (-&25.,3438= 2>(47-.?.*38=)*8=&7'7*8_=*8=(438.)S7&9.438=S;41:9.;*8=8:7=1*8=(-&3,*2*398=)*=5&79*3&.7*8=2>(47-.?.*38= 1.S8=A=1&=2>(4-S9S749745-.*=3*=8439=54:7=1.389&39=+43)S*8=6:*=8:7=)*8=47(-.)S*8=9*25S7S*8`=&1478=6:*=1&= 2&/47.9S=)*8=2>(4-S9S749745-*8=8439=9745.(&1*8_=*=97&;&.1=S9:).*=)*8=,*37*8=2>(4-S9S749745-*8=A=1&=+4.8= 9*25S7S8=*9=9745.(&:=`=)&38=)*:==84:8897.':8`=1*8=)*7;.1.S*8= 5.54,.:2 =*9=1*8=3*499.S*8= *5-&1&39-*7& =*9= 5->1147(-.8 `= &+.3= )*= )S9*72.3*7= 1.)*39.9S`=1S(414,.*= *9= 1&= 85S(.+.(.9S= )*= 1*:78= 5&79*3&.7*8= 2>(47-.?.*38= )&38=:3=2U2*=(439*=9*=5->14,S3S9.6:*_=#&3).8=6:*=1*8=)*7;.1.S*8=3439=5&8=1*8=2U2*8=5&79*3&.7*8=8*143= 1&= 1&9.9:)*`= 1*8= )*499.S*8= 9745.(&1*8= 439= )*8= -9*8= 8.2.1&.7*8= &:== )*499.S*8= 9*25S7S*8_= *5*3)&39`= 1*8= )*499.S*8=9745.(&1*8=7S;T1*39=)*8=&884(.&9.438=343=85S(.+.6:*8`=+&.9=34:;*&:=54:7=1*8=2>(4-S9S749745-*8`= 6:.=84:1.,3*=:3=+43(9.433*2*39=).++S7*39=)*=1&=2>(4-S9S749745-.*=)&38=(*8=S(48>89T2*8_=,&7=&.11*:78`=1*8= )*499.S*8= 57S8*39*39= )*8= *85T(*8= (-14745->11.*33*8= 41= &55&7&.88*39= 543(9:*11*2*39= = )*8= .3).;.):8= (4251T9*2*39=&1'.348=*9=;.&'1*8`=8&38=)4:9*8=2>(4-S9S749745-*8_=*=97&;&.1=24397*=6:*=(*=5-S342T3*=*89= 7&7*=*9=(4397*88S1*(9.433S=a=1*:7=':),*9=(&7'43S=7S):.9`=1&=5->8.414,.*=)*=1*:78=+*:.11*8`=*9=1*:7=&'8*3(*=)*= 5.,2*398=1*8=7*3)*39=51:8=8*38.'1*8=&:=897*88=->)7.6:*`=&:==5&9-4,T3*8=*9=&:==-*7'.;47*8`=*9=).2.3:*39= 1*:7= (&5&(.9S= A= +472*7= )*8= +7:.98_= &= (425&7&.843= )*= 1&= ;&1*:7= 8S1*(9.;*= )*8= &1'.348= *9= )*8= .3).;.):8= (-14745->11.*38= )&38= )*:== *3;.7433*2*398= ).++S7*398= 84:1.,3*= 6:*= 1*8= *3;.7433*2*398= 842'7*8= *9= -:2.)*8=8439=51:8=+&;47&'1*8=A=1&=+.=&9.43=)*=1&=2>(4-S9S749745-.*_=*=97&;&.1=5*72*9=343=8*:1*2*39=)*= 2.*:==(433&97*=1S(414,.*=)*=(*8=47(-.)S*8=+47*89.T7*8=(-14745->11.*33*8=2&.8=&:88.=)*=2.*:==(4257*3)7*== 1S;41:9.43=)*8=2>(4-S9S749745-*8=*3=7*1&9.43=&;*(=1*:78=(-&25.,3438=2>(47-.?.*38=*9=8*143=1*8=7S,.438= (1.2&9.6:*8_= .91*=a= == == = '897&(9a= 0>(4-*9*749745-.(= 51&398= &7*= 34385-4948>39-*9.(= 51&398= 9-&9= 7*(*.;*= 9-*.7= 47,&3.(= (&7'43= +742= 2>(477-.?&1=+:3,._=1:7.3,=47(-.)=*;41:9.43`=9-.8=3:97.9.43=&748*=,*=9.2*8_=2;*7>=,7**3=47(-.)=&7*=247*=47= 1*88= 85*(.+.(&11>= &884(.&9*)= <.9-= 2>(477-.?&1= +:3,.`= 034<3= &8= 8&574'*8`= <-*7*&8= 2>(4-*9*749745-.(= 85*(.*8= &7*= 85*(.+.(&11>= &884(.&9*)= <.9-= 97**= 2>(477-.?&1= +:3,._= 539*757*9&9.438= &'4:9= +:3,&1= 5&793*7= (-&3,*8= &884(.&9*)= <.9-= 2>(4-S9S749745-.*= *;41:9.43= <*7*= 2&.31>= '&8*)= 43= 9*25*7&9*= 2>(4-*9*749745-.(=47(-.)8`=<-.1*=2489=4+=2>(4-S9S749745-*8=4((:7=.3=9745.(&1=7*,.438_=#-.8=<470=89:).*8= 47(-.)= ,*3*7&= 4((:77.3,= .3= '49-= 9*25*7&9*= &3)= 9745.(&1= *(48>89*28= .3= 9<4= 8:'897.'*a= )*7;.1.*&*= 5.54,.:2 =&3)=)*499.*&*= *5-&1&39-*7& =&3)= 5->1147(-.8 =.3=47)*7=94=)*9*72.3*=9-*=.)*39.9>`=9-*=*(414,>= &3)= 9-*= 85*(.+.(.9>= 4+= 9-*.7= +:3,&1= 5&793*78= .3= &= 3&774<= 5->14,*3*9.(&1= '&(0,74:3)_= 6-.1*= )*7;.1.*&*= 57*8*39= +:3,&1= 5&793*78= ).++*7*3(*8= '*9<**3= 9*25*7&9*= &3)= 9745.(&1= *(48>89*28`= )*499.*&*= &7*= &1<&>8= &884(.&9*)= 94= 97**= 2>(477-.?&1= +:3,._= 047*4;*7`= 842*= )*499.*&*= ,7**3= 85*(.*8= 574):(*= 5:3(9:&1= 2>(4-*9*749745-.(=.3).;.):&18=9-&9=8:7;.;*_=#-*8*=&1'.34=.3).;.):&18=&7*=&1<&>8=7&7*=&3)=8**2=94=-&;*=&= 14<*7=+.93*88`=):*=94=&=14<*7=(&7'43=':),*9`=9-*.7=1*&+=5->8.414,>`=&3)=9-*.7=14<=5.,2*39=(439*39_=711=9-*8*= +&(9478=2&0*=9-*2=247*=8:8(*59.'1*=94=5&9-4,*38`=-*7'.;47*8=&3)=1.,-9=47=<&9*7=897*88*8`=&3)=7*):(*=9-*.7= 8**)=8*9_=425&7.843=4+=9-*.7=+.93*88=.3=9<4=).++*7*39=*3;.7432*398=:3)*71.3*=9-&9=9-*>=8:7;.;*='*99*7=.3= )&70= &3)= -:2.)= +47*89_= #-*8*= 3*<= *1*2*398= -*15= :3)*789&3).3,= *(414,.(&1= (43897&.398= 4;*7= 2>(4-*9*749745->= *;41:9.43= .3= ).++*7*39= *(48>89*28`= &3)= 5*72.9= 94= 034<= 247*= &'4:9= +47*898)<*11.3,= 47(-.)8=*(414,>_= .8(.51.3*=a= 8.414,.*=)*8=545:1&9.438`=2(414,.*=*9=2;41:9.43= 498= (1*+8= a= 2>(4-S9S749745-.*`= 2.=49745-.*`= 47(-.)&(*&*`= (-&25.,3438= *(942>(47-.?.*38`= +1:== )*= (&7'43*`=&1'.3.82*`=7S8*&:==2>(S1.*38=(422:38`=(422:3&:9S8`=(48S;41:9.43_= = &'47&94.7*=)&((:*.1=a= *397*=)S(414,.*=+43(9.433*11*=*9=S;41:9.;*`=)9:`=+3+3`=79*=)*=0*3)*`=-.,3-=04395*11.*7`=?7&3(*_=