GENETIC \!4RIATION AND INTER-TRAIT CORRE1,ATIONS IN ELrCAL)II'TLiS GLOBULUS BASE POPULXIION TRIALS IN ARGENTINA

Gusta~o A. Lopez'.', Bradley RI. Potts', Gregory K Dutkowski', Luis ..\. ;\piolaza ' & Pedro 1.:. Gelid','

I ' Coopci-aiivc Rcscoi-cli Ccliirc for Sustliin;,hIc Produciion 1:ili-csiry :rnd Scliool oS P l ~ ~ n i S~.ience. Ilnivcisity oSilis11inni:i. GPO Box 251-55. Hobart. 7001, 'S:ism;ii~i:l. ~\iisIrnlia.

' Pcrinancni adilress: liistitiiio Nacii,ilaI dc T C C I I O ~ I I ~ ~ J ~\g~ope~i iar ia (INl-A) ClliN. Lo? Reseros ) L3s Cah~iiias slil. I51 1. Castc1:ii- ( l 3 . \ ) , hrgentiiin.

" l'royccio Fiiicsi:ii de Desarrolln. (S\GPyh-BIRFj, c/o Los Resciiis y 1.3s Cabaiias sin, 1712. Ca\ielar I13.r\.) Arzcnlina. Ci>rics1ioiidin:l auihiii-: 13r:idiey hl. Piitis, Coopeiaiive Rcsc;ii-ch Ccnti-c Sor Susiainable Psiiduciion i'ori'siry and Schoiil oS

Plant Scicocc, University oll:T,~sinaiiia. GPO liox 251-55, Iliihlii-I, 7001. l:imania. A~~stralia. E-hlail: 13.hl.l'oils@uias.cdu.i1i1

Gcliclic vai.i:itioii and inter-traii cossclatioiis ucrc uxemincd in n h:isc jpo1x'laiion ni Ciri~ii1~ptrr.s ylohi~lirs csi:iblislicil on ioiir siies iii ,\sgciiiina. Tlic gcnciic m:,tcii:il incliidcd 21.3 open ji~~llinalcil sced 101s Sti,m I I Aus1rali:ln localities :ind 52 cipeii pollinated and ciinil-01 ~pi,llinaicil sccd lots 11:som EUI.<I~C:III iind Sotiill Ai~;crica~i l:incl i-aucs. Suri.ii:il. gi-o\i,il~. hark tiiickness. ircc liil-ln, iiansiiion lo ndiilt Soli;igc ;and Pilod)n penclraiioii wcrc :isscsscd. oi diiScl-ciit agcs iip !<>i-yei~s . Average singlc siic iiidiuidu;il narrow-sense 1iciii:iliililics ucrc loiv il~i hi-l:ing(0.03), survi\,iiI (0.05)and S~!rm(O.l0); iniermcdiatc iorgiou.ili (0.27) andsclativc b:iihIhickncss (0 .32): and high Sot Pilodyn [iencifiition ((1.18) :i~id il-:iiisiiion to adiili ioliagc (OhO). 'l'licrc \?,is sisoiig posilivc "cclic correlaiion lhcti$~ccn ilic same lrail inc;isiircd 31 diili.icni siics (> 0.70) and agcs 1>0.64). In gciicial, !lie gclictic coii~l:iiioiis hciii,ccii gmivlh :ind oilier traits ii,cic no! si:itis~icaiiy rigniSicnr11 (Pilodyii pcni-il-aiion -0.0;. hi-I: 0.06. proportion oi;iduli ii1li:ig~ 0.06. bark ihickiiess-0.02). C ~ ~ ~ r ~ l i i i i ~ i i s ai1i01igSI s~ihra~ccliUcis U.CI.C gcnilr3Il)' consisicni with !lie S:iiiiily wiihiii suhi-.ice genelic rorreiaiions. biki lo somc enteiii lower and ICSS sig1iiSi~31ii. Oilier csrimatcs oSgciieiic parutncici-s rcjxliicil Sor JJ. ~lohii1,i.s were ~ C I I C K I I I ~ COIIS~SICIII \villi !he prcscnl study. The limitations oi l ii;iranietcr cstiinarcs (ioin opcn pollinated progeny trials arc ~Iisc~isse~l.

Kcy words: Eirci,I?,~riir filohrilii.~, hciii~ihiliiy, gcneiic cosrclsiion, gcnctic ~i~.iriiiilii. gcnoiype x eiiiii-oninciii inici-nciiiin. iicc hrccding.

INTRODUCTION

Ericii1y~~1rr.s ~ I ~ l ~ i ~ l r i . ~ is 3 kiscsi tree i i i l l i~c ti1 s ~ u t h - eastcsii ~\iistsali; ian~l vai-iously ~ i v c i i specific iBi10i)K- r i t 20011) o r siih-specific sialus (E . glo1~11l1i.s ssp. globr11rr.s; KIIIKI'..YTI~ICK 1975). 1i is the prciilics piiip- u t ~ o d O i c ~ i l ~ p r i i s sl?ccics anti is prowl> iir planlalions in inaiiy icinjicraic coiintrics :~s(juilti lhc W O I - I ~ (EI.I)IIII>GE ei ill. 1'143). It is i i i vasiotls s!:igcs ol'doiiiesticatioii in Inaiiy counu-ics sucli as Australia (TIBBITS el iil. 1997). Clliic iS,ANllt:EI,\ & GIIII:I:IN 2001: GUTII-III<EZ el (I/. 2001), China (ZANG et (11, 1995). Posiugal (i\1cl\uio p t

(11. 19'17), Spniii (Soit1.A & BI:)I?I~..\LIIO 1998) :111d Uruguay (B.AL~,IELLI el oi. 200l ) . Gcnctic parainetci-s arc ril~~uiscil in optilnisc hsccding and dcployniciii sisatcgics. 3s well a s cstiinatc hiccdiiig v;iiucs and failis l io in sclcciioii ( W I I I T E 1996: D o i t i w ~ i ~ o 2001). Tii dcicrii?iiic the rcsponsc lo sclcciioii. il is iinpoi-inlit to

"iveil hiiinv Inow much of ilic pl ic i i~~iypic mi-i~itioii in a ,'

lsait is uiidcr penciic coiiirol (hcsiinhiliiy). li is also iiiilxxlani io undcrsiniid the cxlciii ti1 which illc cs[iscs- sioii o i ilic gcnciic vai-iatiuii is sliible iicross diii'cl-ciri ages ( a ~ c lo age gcnciic corrclniioiis) aiid diSScsci?c sites (as a11 indicatiuii of pcnotype s ciivironinciit inter-ac- tioil). as wcll 35 ilic geiictic cosrcl;iiic~ns existing ainr1np" cliifcrciil trails Lo pscdici !lie i-cspoinsc 10

sclcctioii i i i a muliiv;iriatc scirsc (F i r . cosEi t & M.ACK- .AY 19%).

.\.lost stiidies o i gcnciic pasamclcss i i i E. :.iohlr/ll.s lh;lvc I'i~cuscd on gsowlli (c.g. I'WI-~IS & i o i c o i ~ 19941~; B o i t i < . ~ u i o ei (11. 1995: MOI>GE er 01. 1996: BAI.;\IELLI c,t ill. 2(101) and iis associalion wit11 a Sew hey ii-aiis siicl? I I siii-viva1 ( C I I A I I I I ~ E I I S pti i l . 1996). \vood dciisii)' (Boi<l<il,iIo el 01. l992h: M,\cDo~..41.u el ill, 1997: f v l u u ~ ~ t i & Rr\\i.hiox~) 2000). i'Ii)\vcriiig ~)I.ccoc~~). l ~ i l ~ \ h l l 3 ~ l ~ S e l i i/. 1'197). forill (~OI.KEIt ef o/. 1990) niid vcgiliniivc phase cliaiigc ( I i ~ ~ u z ~ i er iil. 19114; Jc1lil~..\N <,I ril. 201)O). Thcsc siiiilics dil'lcr in aiiaiyii~;ii

Iechniiluc. statislical inodcl. genetic inaici-ial siiidicd niidlor lcsl ciivironincnl.

Tlie present siiiily reporis tlic gciictic paramcicrs i j r lhc firs1 pcdigrced E. giohif1ii.s irials grown i n Argeii- tiiia. Tlicsc ti-inls combine lioili n;itivc stand seed lois froti? ~\usii.;ili;i ;tiid l;iiid I-ace inaicrial llioin around ilic world lo construct a Iaqc hasc population liir breeding rliis species in Argentina (Loi'irz er cil. 200113). \Vc cxail~ined the level of genetic conirol ;tad genetic correlalions for eight traits covering growili. wood deiisiiy, iorm, hark iiiickncss and vegctaiivc phase cl i~lngt The csprcssion of gciiciic vari;tiioii i n ilicse trails is cxamineil acnrss kiur silcs, its \%:ell as llie cll:ingc with ;!gc up io 4-years.

hl,iTEIII.ILS AND RIIS?'HODS

Plant nmleriai and trials

'Tlie four sites siudicii arc 1oc;iIcd i n llic Bucnos Aircs Province. Argentina, \viihin the Iraditioiial E. g/ohil/if.s pliiiiting zonc(dcSincd ill ~~OPEzeiiii. 200lc). Tlie lrials arc iilenlified ;is BALC (h i , 37"45' S 111ng. 58" 17' W). BOSC (lai. 38"39'S long, 59"Od' W); MANU (121.37" 53' S long. 59' 56' W) atid VOCA (la[. 38" 28' S long. 59" Oh' W) ant1 vary i i i avcr;rgc annual miniall froin 850 to 1008 inin and average aniiual rniniinuin tcmpcfii- turc from 5 . 8 " to 3 . 3 "C (see l A ~ ~ ~ ~ e i a/. 2001h). The trials included a tolal o i 14,925 irces from276 seed 101s. T\vu linndrcd and iwcniy thrcc open pollinalcd (OP) lhniilics and one hulk collcciion wcrc Srom 1 I inativc stand localities in Australiaand 52 seed 101s wcrc 1.1-orn land races fso~nPonugaI. Spain. Chile and Argcn- lina. The land race seed 101s iircliidcd 37 OP fai-iilics and 10 coniroi pollinaied (CP) lamilics. as well as 5 bulk collcclions o l varying lcvcls of genetic improve- mcnt. Tlic nu~nhcr of I;i~iiiIics ( i .c seed lots) licr trial ranged from 220 lo 275. Full delails of the disirihuiion of falnilics across collcclion localilies are given in LOPEZ et 01. (200 I h).

Tlic trials comprised 15 rcplicaics per IBmily of siiiglc-tree-plots. Families were arranged i n sets o i 2 0 lo 25. based 011 their gcirgr;ipl?ic provenance. Scis \vci-e mndoinly al l~~caled 1 0 apositiun in ihc trial and Linilics wiihiii each sci ivcrc randonily arranged in each repli- c;iie of tlic scl. Sp:icing was 3 in? x 3 m iii ail irials.

Mcasuremcnls were madc of growtli. aduit foliage. Pilodyn pcnctration. l3:irk ihickiiess. forin and survival. Sccm ilianlcicr (cm) over bark was assessed at hreal

lieiglit (1.3 171) at 2 (UB112). 3 (DBH3) nlidd (DBHJ) ycai-s aftcl- plantins. Toial hciglil jciii) was ii~casured at I (HTI ) and 1 (Kr2) years. Tlic proportion of ;idiilt fi~liagc was asscsscd at agc 2 (ADFO). u.hilc ircc i~ r i i i (FORh.1; 1 worsi - 4 hesij was asscsscd ;it ago 3 (MANU and VOC/\) or 4 (BALC aiiil BOSC). These last two variables !vcrc mciisui-cd on an oriicrcd 4-poiill scale [dclails in L o i ~ r z er (11. 2001 hi. Bark ihickncss (BTIII) :iiid PiIodynl3ciietration (PILO) i\~ercmcasurcd (mm) only in B A I L and BOSC ni :igc 4. Tlic ~aiialile BARK useil for analysis is the pcrceniagc oil1Bl-14 ili;ti was cliectivcly hark for cacli iiidi\,idual ircc. TIic 13rcseiicc of forks a1 age 2 (FQIIK) \\a:, asscsscd iii ilic sainc 2 11-ials apgroxiiiiaicly 2 ini~iiilis aflcr lighi 1i;iil

daiiiagc. Survival (SUIZV) w;ts dctcrmiiicd htiscd oil ilic oldcsl asscssinciit lor each trial (i.c. ;igc 4 &IF BtiLC and BOSC and age 3 ior MANU and V0C.A).

The geiieral set of data used for :iiiaIysis was lroiii all tliose 11-ces Il1ai wei-cali~e :ti ine;isurciiii'iii: I I O W C ~ C ~ . isnlakd ou1Iic1-s (usii:iIly runts) were reiceled. At MANU approuiinatcly 25 5:. of trees suffered damage hy coivs a1 an early age and \\,ci-c cscludcd li-oin ihc analysis.

Analyses

Forcacli trial, ail incolnplcic hlock design was imposed (I i'osferiol-i (see ERICSSON 1997; FI! ei it/. 1999). Tho i~icoinplctc hlocksize used forcach trial was ihai wliicli inasiiniscd tl?c ~n<jdel lihcliliotrd for the oldest diaiiieier Incasurcmcnl ( D U T K O ~ S I ~ I e i it/. , 211112). x~ria~icc com13onenis lor cacli trail wcrc c<i,itiaicd with iiii

individual ircc~nixcd iia~del. using rcslricicd m, i i ~ ~ i i i u ~ i i . '

likciihood i~?iplcmcnied will? ASlic~ni ( G I L ~ I O U I < ef iii. 2001 j. The inodcl filled was:

whcre: y is thc vecior of individual tree data: b is ilie \'cctor fol- ail ilie fixed effects io\,ernll iiicaii, si~bracc ;is dcfiiied by the clussificalion o i D u ~ i t o u ' s ~ r & Poi'is (1990) or Iiind I-iicc (I2oi,~z 01. 200113). ~ind polli~ii~. tion type-opcn~?oIli~ialcdorcoi?~rol po1Iiii:licd): a is 11?c \rector o l itnobsei-vahie additive gciietic cflSccts of iiidividual i r c c ; c is lhc vccior Iiir tiie inndoiii effects of i!?e iiico~iil?!cic blocks and e is the vcciot- of ~.esidu- als. XI Z,, aiid Z,. ;Ire incidence inntrices I-cI;iliiig ilic ohscrvalions ~ L I ilie Sised and i-andi~in ciiccls in ilic modcl. Model I was filled for both uiiivariate and hivariaic analyses and in holh cases; ilie espectcd mcair and va~.iaiices of tlic pai-amcicrs arc:

u ~ l i e r c V = % , , G , Z ' , + Z , G , % ' , + R , G , = A @ G , , ivhcrc ii is the nuincraior rclationsliip matrix, G,, is tlie additive ge~ietic covariance matrix, and @ is the KI-onechcr product, G,= 1 @ G,.,, wlicrc I is ail idcntiiy matrix, G,, is the inconipletc block covarimcc mauis and R= R, o I wlicrc R, is LIic treit residual coval-iancc matrix. For llic univariatc case, soinc of these i~atriccs (c.g. G,,. R,) coilapse to scalars (c.g. ot. 0;). ~ h c additive relationship inatrix ii was inodificd lo account S I I ~ ail assuiiied 30% sclii~ig mlc, which increases ihc sih additive coefiicicnt o l relatedness (r) from 0.25 to 0.4 and tlic parent-i~lfspriiig relntcdiicss from 0.25 to 0.4 as \veil (DUTKOI~SKI ef 01. 2001). Tlic Sew hulk sccd saiiiplcs wcrc iiicltidcd in tlic rclatioiiship matrix wiili missing inalc and kiiralc pedigree inihrmetion. A binomial ~nodcl was fitted to prescnce/abscnce traits (SURV, FORK) with n probit transforination. The significance of random effects \%,as teslcd usi~ig a likelihood ratio icsc (SE-IRLE 1971).

To compare the absolute levcis o l additive gc~ictic ~arialioii a c r ~ s s traits, the cocSiicients of additive genetic variaiice werc calcuintcd as:

0 cv,, - I00 4 Y

131

\vlicrc: o,, is tlic witliiii suhracc additive genetic sian- dai-ddcviatio~i calculatcdlio~nthcui~ivarintcmodel and ,7 is liic popuilliion mean (aflcr IHOULE 1902). Single sitc, nnrx~w-scnsc hcritnhilitics (11') wcrccrilculated as:

ivlicrc: o: is !lie addiiirc gcnciic variance witliiii snbl-aces and G~: is tliccnori.ai-iancccomponcnl in 121. Tlic crror icrin iiicludcs the spcciiic comhiniiig ciicci l io~i i ilic few conii-oiled crossed fiuiiilics. Tlic sia11da1-d crrors ofcsliinalcs wcrc calculaicd hy ASlieinl lioin ilic avci-agc inSonnation iiralrix. tisilig a sia~idard truncated Taylor series approximaiion (GILIIIOUR et 01. 2001). Geiietic diffcrciiccs hct\vccn suhraccs werc tested with llie F-statisiics using an crror degree of freedoin del-ived li-orn tlic family wiihin subrace teri~i. Subracc lcasi sqiiarc means aiid their skinditrd zrir~rs \\,ere also estiinaied iii ASRcinl. Pwrson's correlation coefficients aiiioiigst tlicsc subraccs least squar-c niczins wci-c cstiinatcd using ilie PROC CORR procedure i n S:ZS

(version 8). Pairwisc gciictic correlations aiid ilicir siaiidanl

errors were csiimatcd using 3 hivai-iatc inodcl ilia1 was cxieiidcd from ilic uiiivariate ~iiodel. \Vhcihci- corrcla- lions were siyiiificantly different fiorn zero ilr not was deierrnincd \villi a lihclihood ratio test.

RESULTS AND DISCUSSION

Expression of genetic effects

The gciictic variation was pariiiioiied iiilc fixed sul~rilcc clfccts and random iidditivc genetic variatioii \vitlhiii suhracc. The estiinatcs pi-cscntcd in illis study arc dominalcd by open pollinated ~iaiivc skilid fai?iiIies (82 %) butalso integrated infoi-mation froini~pcnpoiii~latcd (14%) and control poliiiiatcd (4%) ltiiid race F~i~iilics. All tmiis slio\vcd sigiiiiicant subrace diiicrciitiaiioii. cxccpl survival at PvliiNIJ (Tahlc 1 j. The patterns of differentiation at Ihc subrace icvcl are discusscii hy I.OPEZ er ill. (2001 h), ~hcreforc this pajicr C O C I I S ~ S OII

the genetic val-iation o l iaiiilics witliiii siibraccs. sf. P L I s I I c ~ I I ~ . ' , sig~iifi~aiit I c~e l s o l additivc fciiclic

variation within subi-aces wcrc dctccicci lor all isails cnccl~l survival. ivherc signiiicanl ~'ai?aiioii \\.as oiily dciccled at VOCA (Table 1 ) . f-lo\vcvcr the mrignitudc ofllic addiiivc variailcc was scalc ticpendent aiid ii i the case ol~growtli. incrcascd with ilic incasurcinelii incan. Tlic ability o l trilits to rcspoiid lo selcctio~i caii he compared bytl iccoeff icici i toladdit i \~cgci ict ic \'ariaiicc (CV,<) that lncasurcs lcvcls oladditivc fciictic variation wliilc accounting for scale and size c i i c t s (HOUI-E 1992). The CV, was close to 1054, SIIS most traits. cxccpl for the proportion of adult fc~liagc, iiliicl~ was grcatcr tlian 10% at ell sitcs (Tahic I ) . 111 a rcvicu o i ttic inii~i~iludi. o l tlic c~ieiTiciciil 01' :iiIilitiii. gcrictis \,ariaiir,ii iicrosb a large iiuinber oiqii:iniitativc traits in a variety of Sorest ircc sl?ccics. Cot t~i i . ius (1994) reported incdian v;ilucs for lieigl?t and diainctcr growti? o l 9%. consistcnl \vitli tlie average 111it:iiiicd i i i tlic prcscnl siudy.

The niagiiiiudc of tlic gcnoiypc u ciivironmciit iiitcrilction oil trail cxprcssion ctiii bc gauged hy llic gcnciic coi-rclatio~i bci\t;ccn tlic siinic irLiit inc;isi~rcd ai diilcrciit sitcs. Such corrclations cotil~i be partitiiinctl inu i tlic corrclatioii of siibi-ace ciiccts (r\) aiid ilic corrciaiioii oladditi\~c ,genetic cl 'l~cts u-iiliiii suhrticcs (I-,). The suhrace c ~ r r c l a t i ~ ~ ~ i s across siics werc siroiif aiid positive lhr inosi lriiiis (Tilhlc 3j; iiidicatiiig stal~il- ity ill' subraces periormaiice foi- lhc traits ohscrvcd. Gciictic correlaiio~is within suhi-aces wcrc also liigli (T.ihlc 3). and in gciicral sligliily lhighci- iIi;ii? i i x

sul?rilcc c~~rrelrilioiis.

Table 1. The overall mcans with standard deviation (sd), additive genetic variance (I) with their significance (Sig.) and coefficient ofadditiw geneticvariation (CV,,in %units), aithinsuhrace heritahililies (11') with their standard errors (s.e.) and the F values and significance of the subrace eLfects (F) for the 34 uariahlcs measured across the four trials. Significance of effects arc noted as ns = not significant; * P < 0.05; ** P < 0.01 and *** P < 0.001; NT = could not he tested. "n" represents the nun~ber of data Lhr each trait included in the analysis. The traits NT (height, cn~), DBH (diameter at breast heiphl, cm). PILO (pilodyn penelration, mm), BARK (bark thickness, 'c), FORM (stem form), ADFO (adult foliage, prtrporiian), FORK (presence of fork, binary) and SURV (survival, hinary) are follon,ed by a number 1~11ich represents the age of measurement in years. Note that n in DBH3 at BALC and BOSC are smaller that DBH4 due to exclusion of dala a1 age 3.

Addilive eenelic Heritabilitr Suhrace eSfecl Trait Site I Z &lean s.d. - ---

ol, CY, I,' S. e. F Sie.

HT I

HT2

DBH2

DBH1

DBH3

PlLO

BARK

FORhl

ADFO

FORK

SLRV3

S b R V 1

RALC UOSC

BALC nose VOCA M/\NU

BALC ROSC VOCA MANU

BALC ROSC VOC A MANU

BALC BOSC

BALC BOSC

BALC 13OSC

B/\LC BOSC VOCA MANU

BALC BOSC VOCA MANU

BALC BOSC

VOCA MANU

RALC ROSC

3.5 *** 13.6 Xi"*

6.7 *:i:i

10.0 **/ ,. ,.

15.5 .*:il:

12.4 ***

6.0 *ii

7.7 I:**

9.4 I:**

10.5 **%

4.2 **:t

3.7 *L*

8.5 *** 4.4 i:**

3.0 iiii

2.3

11.6 *** 7.0 *:*a

21.9 *:be

13.5 eii*

*** 12.0 ,. . ..

13.1 *"*

14.4 *** 7.0 a***:

7.8 *** 5.3 *** 10.3 li(*

5.8 l;l i

6.0 a.

4.4 **.? , ... ..

2.3 A

1.6 ins

3.1 *** 2.0

Tahle 2. Individual narrow-sense heritabilits estimates (h' ) I i~r rrorr.111 traits af 13scalvntrrs rlobrrlrrs OIICII oollinated .. - . . familics from single sites or the sites average rvhen multiple sites are reported. The coefficient of relationships (rj used in the eslimalion are riven and the heritabilities have been standardised to a common basis n f r = 0.4 Or' for romoarison. - ,,,,, IVherr r-is not given these are treated as r = 0.4. The traits dhh (diameter at hrcast height), 11t 01cight) and rol (rolumcj are followed by a numher which represents the age of measurement in years. The group effect represents the inclusion (Yes) or not (No) in the statistical model of a fix effect to account for the provenance, subrace, rare or other group effect. Number ofhmilies (No. of families) and number of sites (No. of sites) tested, and type of population (Pop. type) refers to whether lalid race (L) or doniinantly Australian native stand ( N ) families have been tested.

Group No. of I'op. No. of Trait h' I' Country Source ol reference and year

effect families tvoe siles

dbli ? dbh2 dhh3 dhhi dhlil db114 iibil4 dhh4 dblii dbh4 dhh5 dbh5 dhh5 dbh6

No Yes Yes No No Yes No Yes Yes Yes Yes No Yes Yes

L X N 5 + N I N I N 3 N 4 N I N I4 N 5 U 5 N 5 N 14 L N 3 N I

Poiluglil Chilc Chile y . lugiiay i\usiralia, VIC Chile filusl~tlia Ausir~lia, TAS Ausi~ilia, TAS Australia. TAS S[i:iin Uruguay Australia Australia. TAS

ARhUJO a1 a/ . (1997) ~I'INLh e l n/. (1994) lI~lXZ,\ el a/. (1'1941 HAIAIEI.I.1 el it/. (2lIOl) L~~)OI..\STON CI ZZI, (I 9'1 I ) II'INZA el iil. (I 991) l3ill<R,\l.HO <t/. (I 995) ROIIR~II.HO X POTI'S (19'16) I\.l,\cDou..\i.i> e i a/. ( 1 997) DU'TKO\VSKI el nl. (1997) Solti:\ c i ',I. ( 1 997) HA1.nilii.l.i el a/. (200l j \lUNTiRl& RhYhlONil (2001) VoI.KliR <,I 01. I 19901

lhr '!

Ill 1 1112 lh12 1112 1713 I h l i

lht4 I114 I114 1115 l h l i ill6 1116 hi8 Ill')

No No No Yes No Yes No Yes Ycs No Ycs No Yes No No No

No Yes Yes No

Yes Yes - No Ycs -

560 L X N 5 + 20 L I 20 N 4

224 N I 20 L I

223 N I 89 L 3

224 Y I 5'14 N 5

20 L 1 260 N I4

89 L 3 45 N I 27 L 1 33 L X N I 27 L I p~~~-~~~ ~~ ~ ..~,,. . . .

560 L & N 5 + 23 N 5

X N 5 89 I.

223 IU I 539 N 1 653 L & N - 89 1. 3 35 Y 1 - - I

Purlugal Porlu~al Aosiialia, VIC Chilc Poi-Ilig"1 Chile IJrugiiay Chile Australia, TAS Poi tu~al S p i n Urugi~ay Ausiralia, TAS POI~IU~:IJ P111.1ugul Pill.i"fal . ~. .~~~~~~

Portugal Australia Australia Uruguay Chili. Ausil-uli:~, TAS Cliilc Uruguay Ausii-:ilia. 'I'AS Pllrl~ical

AI<AUK) (,I a/. (1997) Roi<i<ni.ilo el 01, (1992) Wool.~sTou el 01. ( I 99 I ) ll~lNii:\ <!i <,I. ( I 994) HoRRni.Ho e l nl. (1992) IPiNzn e i 01. (1'194) Hnl.h~il~Ll.1 el i d / . (2001) 11'iNiL2\ el 01. ( I 994) Hoi<a:\i.iio el 01. (1 996) l3Ol<l<~\I.l+O <,I (!I. (I 992) SOKI:\ el 01. (1997) R:\L~II?I.I.I e i nl. (2001) V~)I .KI~I< (!I 01, ( 1990) I ~ o I ~ I ~ : \ I , I ~ o c i <,I. I 1992) D0i~l t~l .110 e i 01. ( 1 992) HOl<l<:\l.IiU ti (,I. ( I 992)

I%K,XUI~I e l 01. (19'17) H0uc;ii el 01. i 1996) Houcii el a / . (19'16) ~~\l.hllil.l.l el i d / . (200 I ) II'IN7.A el n/. ( 1 994) POTTS X JORl>r\N I 1994) \IEI<G,\Ic,\ X CIIII:I:IN I 1997) Hhl.hllil.l.l <!I ( 7 1 . (2001 ) VOI.KIrt cl a/ . (I 990) C0T-i-lil<li.l.& Bl<ol.lN (1997)

"I Ileiitabililies esiiinalcd by ihc aullioi-s ibr Auslrnlian native stand familics havc no1 been included i i i tlic a\,i.i;izc prcsciiicd here a tliese \\,ere liigliiy inilalcd duc lo incliisioii o i 1ar.i~ provenance eiScuts i n the diffci-cnce beiivecn ibmilics.

Table 3. Across site correlations between subrace least square means and their significance from zero. Thc significances are noted as ns = not signilicant; * P < 0.05; ** P < 0.01 and *** P < 0.001. The traits HT (heieht, cm), DRH (diameter

represents the age of measurement in years

Trait Site BOSC VOCA hlANU

BALC

BALC BOSC VOCA

BALC HOSC VOCA

HALC BOSC VOC,l

DBll4

ADFO

BALC

HALC 13OSC VOCA

FORM BALC BOSC VOCA

PILO

BARK

FORK

BALC

BAl.,C

High iutra-population genetic corl-elations (r8) a-e often associated with high inter-population (e.g. I;)

genetic correlations (ZENG 1987) as in tlic prcsent casc. However, this is not always tlic case i n E. globirlirs (POTTs &JORDAN 1994a; JOIIIIAN et ni. 2000) and to some extent tlic pattcrns of subrace correlation will dcpcnd upon the pattern of subrace sampling. In conti-ast, thc gcnctic corrclations within subraces al-e likely to be more stable and reflect plciotropy. Tlic pooled estimates across dilfercnt subraccs would argue against an elfcot oilinkagc. In the present casc. mea- surements o l thc same trait at different sitcs were genetically correlated at a level usually greater than 0.7 (Table 4) and, at least for ]nost growtli rncasurcmcnts. this was also the casc at t l~c suhracc lcvcl. Thcsc results would a]-guc that thc same genes arc being cxprcsscd at tlic dil'fcrent siles and that frolna breeding pcrspcctivc a single bl-eeding population would hc suitahlc for the range of sites tcstcd.

Growth

Fast growth is one of the major objectives of cucalypt breeding programs around the world (GIIEI\\'ES ef a/. 1997; BORRALHO 2001). Growth was under moderate genetic control in our trials. In general. single silc heritability cstiinatcs for growth traits rallgedfroi1l0.09 to 0.36 (Tahlc 1 ).The average heritability offour single trial estimates for diameter across ages was 0.25. Estimates from MANU were atypically low (0.09 to 0.12). but this poor expression ofgcncIic variation for growth within sub]-aces at tliis site was not reflected in the cxprcssion of subracc diffcrcnccs. Heritabilities Sor height ranged Croln 0. I 1 to 0.36.

Acomparisonwitlihei-itability cstimatesrcported in the literature is complicarcd by thc dillcrcnccs in cnvi- ronmcnt, methodology of analysis (c.g. whether group effects are excluded or not), age of evaluation and the coefficient of relatedness ( r ) used. Tahlc 2 lists the herilability Sol- growth traits reported in othel- OP

Table 4. Across site ge~ictic correlations t standard errors derived Sroni pairwise analyses. All correlations mere signilic;inlly different from zero (';** P < 0.001). The traits IIT (height, cm), DBH (diameter at breast hrighi, ern). PII,O (pilndyn penetration, tnni): BARK (hark thickness, proportion), FORM (stem form), ADFO (adult foliage proportion), FORK (presence of fork, binary) and SURV (survival, binary) are Solla\r.cd by a number which represents the age of measurement in years.

Trail Site HOSC VOCh MANU

DBHI

ADFO

PlLO

BARK

FORK -

13,tLC

B,\LC BOSC VOCA

RALC 130SC

VOCA

BALC BOSC VOCA

RALC

BALC ROSC

VOC/\

BA1.C DOSC VOCA

13AI.C

BALC

BALC

sludics o f E g iuh~~ius ailcrslandardisingthccoeiiicient o l rclaiionsliip 111 0.4 lo allow a heltcr comparison hciweell slutlics. Ouraveragel?eritabilily (0.25) Ihrlrcc diamclcr across sites aiid ages was close to tlic hcrilahi- lilies rcportcd for most Auslralian trials (BOI<KALHO el al. i995,0.24; MACDONALII eta!. 1997.0.26: MUNEIU & ~ ~ A Y ~ I O N ~ ) 2000, 0.21). Other studies (WOOLASTON ct al. 1991; I P I ~ A et nl. 1994: A J ~ A U J O er CI!. 1997: SORI,\ er ill. 1997) reporled much lower hcritahiiities for diamctcr (0. I I to 0.15). Tlic average liarrow-sense liesitability for diamclcr, wciglitcd hy the iiu~nbc~- of farniiics i~icludcd i n each of the studies rcvicwcd i n Tahlc 2 (iiicluding our study) is 0.21. This value could suSl.icc >IS s1and;lrd valuc for llic OP Iicrilahilily of diiimctc~- growti1 in E, glohiil~is.

T l ~ c average heritabilily ior height (0.23) across the iour Argentinian trials was similar to tliat for diarnctcr and cquivalcnt lo ihc weighted average o l rcportcd csiiiiiatcs ior hciglil(0.23) given i n Tahlc 2. While lrec volume was not suidicd in the prcsenl case, tlie weight-

ed average heritability for volume (11' = 0.28) in Tablc 2 was o i the same magliitude as lliosc rcporlcd f ro~n other E~ic~i!\prirs spccics, such as E. iiropiiyl!u and E. grzi,iiiis (/I' = 0.31 and 0.29, respectively) (reviewed in REZENI~E & 13E RESENUE 2000).

Thcre was a consistent increase in heritahilily eslimates for diarnctcr with age in all four Argcntiniaii trials (Tablc 2). This trend has hccn previously rc- ported ill olher iorest tree spccics (as rcvicwed in WEI &BORIIALHO 1996), iiicluding E. gIohii11i.s (BORR\LHO et 01. i992a). Sizc-dcpcndcnl mortalil). (CHAMRERS t't a!. 1996) or seicclivc tliinning lnap l~iash such age trends in tliccxprcssion ol' genetic varialion hy rcduciog ilic amouni o f genetic variatioli (WE1 & BOI<l<ALI~IO 1998). Howcvcr, this is unlikely lo be i~nporlanl in llic present case. Mortality was low (Tablc I) and when accounted foi- using a multivariate analysis combining fourth ycar diamctcr will? iirst ycar hcighi meastire- mcnts, tllc heritability o l Sourili ycar diainetcr only inargiiraily increased over the univariatc csiimale 1c.g.

0 I P U B L I S H E R S 223

G. A. LOPEZ ETAL.: GENETIC VARL4TION IN E U C ~ I / ~ Y ~ ' ~ ' U . S GI.OBI/I.US

, - , ,, - , cm), PILO (piladyn penetration,mm),HARK (bark thickness, proportion),FORhl (stem form). ADFO (adult foliage, %), FORK (presence of fork, binary) and SURV (survival, binary) are fi~llon~ed hy a number which represents the age of measurement in years.

Site Trait HTZ DBHZ DBH.1 DRHl PILO BARK FORRl ,\DFO FORK

BALC HTI HT2 DBH2 DBH3 DBI1-l PlLO BARK FORM ADFO

BOSC HTI HT2

PiLO BARK FORM ADFO

VOCA HT2 Dl3112 DBH3 FORM SURV3

MANU HT2 DBH? D13H3 FORM

0.80 ";* 0.65" 0.70" 0.69' -0.67" 0.77*'; 0.04ns 0.2911s 0.81 *= 0.70 " 0.71 * -0.66 * 0.68 * 0.1 I ns 0.20 ns

0.92 ';** 0.92 *"'; -0.53 ns 0.74 i;' -0.19 ns -0.19 !IS 0.96 ""': -0.54 ns 0.84 ** -0.06 ins -0.24 ns

-0.62 " 0.79 "; -0.08 ins -0.22 ns -0.71 * -0.50 ns 0.00 ns

0 . i s -0.26 os 0.21 ns

0.87 ":be 0.80 *;; 0.74 ** 0.46 ns -0.50 ns 0.69 * -0.41 ns 0.42 ns 0.90 a ~ k * 0.78 *" 0.76 ** -0.54 ,IS 0.62 " -0.22 ns 0.34 ns

",'1[, ""8: 0.88 ""* -0.40 ns 0.62 " -0.46 ns 0.09 ns 0.72 * -0.38 ns 0.64 * -0.58 ns -0.20 ns

-0.33 ns 0.44 11s -0.24 ns -0.03 ins -0.66 "; -0.23 ns -0.25 ns

0 . 1 9 ins 0.01 ns 0.24 ns

ac B A I L 0.33 and 0.15 and BOSC, 0.35 and 0.36, respectively).

Thcrc were vcry strong age-age correlatio~is in tlie expression oC genetic variation both within (rx; Table

6) and betweell subraces (r,; Table 5). As would be expected, tlie magnitude of this gcnctic corrclalion declined with increasing age interval. Howcver, even at the greatest intcrval, tlic corrclatio~is were still highly significant and the genetic corl-elations ( 1 ; ) between 1 year height and the I year diamctci wcrc greater than 0.6. Wliiic subrace cllects were less strongly correlated across ages than the genetic effects within subraces, in most cases these correlations were high and slatistically significant. While these trials are still relatively young. a strong genetic correlation (r, >0.7) between hciglit at age 2 years and later age scctional arca ~ncasurcincnts (8-18 years) has been reported by BORRALIHO el ill. ( 1 992a) for E. glob~rl~rs. Sectional arca estimates at age 4 ycars wcrcaiso highly correlated (r,>0.95) with later

-0.06 ns 0.50 11s -0.22 ns 0.41 ns -0.22 ns 0.37 ns

0.00 ns 0.20 11s 0.14 ins

0.18 11s -0.34 ns -0. I3 ns

0.00 ns -0.02 11s

0.20 ns 0 . 0 5 ns -0.3 1 os

0.26 ins

0.54 11s 0.49 ns 0.57 ns 0.33 ns 0.45 ins

-0.23 ns 0.28 ns

-0.49 ns 0.48 ns

age nicasurclnenls taken close lo rotation age of he- tween 8 and I8 ycars. Thcscrcsuits arzuc that selection for later age growth could occur as early as 2 ycars, hut BORRALHO eta/. (l992a) argue that the optimal age for selection is four ycars or, ii cxprcsscd in terms of avcragc tree size. Xin of heiglit.

Despite tlie environmental differences between trials (LOPEZ et a/. 2001 b), growlln was strongly corre- lalcd aci-oss sites for all pair-wise coinparisons ( I : > 0.78). Growth on the dirl'ercnt sites call thus be consid- ered the same trait because family pcri~rmance was stahlcand thcgcnotypc by cnvironmcnt interaction was vcry sinall (BURIION 1977). Comparable witliin race genetic correlations across sites were reported by M A C D O U A L I ~ ~ ~ C I ~ . (1997) for4 ycar-old diaiiieter oSE. glnh~r1i1.r across five sites in Tas~nania(averaged 0.80 cC 0.91 for comparable ages in the present study). In another Australian study, M U N E I ~ I & RAYMOND (2000) reported similar gcnciic corrclalion bctwcen two sites

G. A. L.Ot'EZ IT.4L.: GENETIC \'41tli\'~ON IN EUCiLYPTLfS C;i.~ilL'l.Lr.Y

in \Ticloria (0.76) hiit growth a1 tlicsc sites was poorly 1997) a~idsliglitly higllcr than tlic average (0.82) across corrclatcd will1 a third sitc ii i Tasrilania (0.12 aiid 0.29). sitc gclietic con-elations rcportcil by .L.lUNEItl R: lii>-- Tliccorrclatioo ofsubraccpcrfor~nanccacross sitcs \\,as hlON11 (2000) lor \>\,el- 70 i;lmilics on thrcc Ausiralian also us~~a l ly strong. but 11i IOU.CI- significance tliaii the lrials. The corrclatio~i hetween suhracc Piiodyii pciic- cenetic correlations (Table 3). In contrast, more vari- lralio~i ~lcross the two Argentinian sites was very high able]-acc corrclatioiis wercrcpoi-tcd by Mi\cDoN..\Loet ( r , = 0.94) MACDONALE ef iii. ( 1 997) also reported 01. (1997). wliicli in one casc was even of different sign liigli correlalioiis iii race pel-iormaiicc across sites for to tlic additive genetic coi-relations within races. sug- Pilodyii penetration (average 0.79). Thcsc regular gcsting ihc Tasmanian silcs wcrc inore divergent than I-cporls of high genetic col-relations i n Pilodyn pcnetra- tliosc in the prcscni study. tion across sites al both llic genetic and suhracelracc

lcvcls clearly indicate little genotype x c~ivironmc~it Survival interactioii occurs 1.~11- wood dciisi:y in E. giohiilris,

consisterit with the iliidings of M U N E I ~ I R: II,\YklONIl There is intcrcst in including survival as a selcciion (2000). crilcrion in E. ~1nhuiu.s breeding prograins whcrc the aim is to rnaxiliiise prodiictivity on a per hectare hasis Bark thickness (CHAAIBERS et (11. 1996: C H A ~ I R E R S & BORRALHO 1997; TOIIO et '11. 1998). This is pmticulmly impel-lnnt Rcl;iiivc hark thickness may ~ io i only aCicctcslim;ltes of iii more cxtrcmc cnvironrne~its where mortality duc lo uiider-hark volume. hut Tilay he o i adapiivc signif- dl-ought or ii-osl may have an eco~ioinical impact. In the cance, affecling the susccplihility or rccovcry of trees present casc, surviual was liigli (>88 %,), and had the liorndamage by pests (c.g. P l ~ o r z i c o i ~ r i i i ~ s ~ ~ r ~ i ~ ~ ~ r r i c f n t i i

lowest Iieriiahility 01 all traits. The heritability o i -SORIA&BOIIII:\LHO 1998; P e ~ a <!ffirilrs - DUTKO\V- survival avcl-ycd less tlian 0.05 and only at one sire SKI&POTTS I~~ '%JOROAN ef~zl. 2002) orcnvironmcn- was this signilicanlly diifereni from zero (VOCA, Val stress (e.g. drought - DUTKOWSKI R: POTTS 1999). 0.13). CIH,\MBERS etai. (1996) rcportcd that heritability Estiinntesollicriiahilily l,rrclaiivc hark thickness were for E ~iohulirs survi\.al ranged li.o~n 0.19 to 0.57 iii significant hut modcralc (0.33 and 0.31) and higlily sitcs in Australia and Ponugal, with sites ]nost affected sig~iilicantdifferences wcrcobserved hetween subl-aces by drought having lower 11~1-itahilities for survival than (Table I). The genetic correlaliori 01 bark lhickiicss those affected hy 1.1-osl. While suhtlc dii'ierences in acl-oss sitcs was liigli hot11 withi11 (0.88) and between survival wcrc dctcctcd hetwccn subraces at all hut oiic (0.83) suhraccs. DU'rKOwSKl ef 01. (1997) and KELLY Argentinian site. Ilie trials were not subject lo major (1997) rcportcd similar estimates of witliiii race licrila- stress during tlie period studied and ge~iclic variation in hilily. The lalter author also notes lliat rclaiivc bark factors aifccti~ig sul-\;ival was poorly cxprcsscd. Iliickness remai~icd fail-ly conspant across ages (4; 5 and

8 ycars oiagc), sites (5 in Tasmaniaarid I in Spain) and Pilodyn penetr a t' ion measurc~neni tcchniquc.

Alter growth. wood density is tlie second most impor- tant sclcction trail included in E, globiiias breeding programs aiiiied at pulpwood production (BOICICALIIO ef oi. 1993; BORKAL110 2001). W ~ o d density is often measured illdirectly in such prograins using Pilodyn peiietratio~i due to tllc io\vci- assessmen1 cosl. Pilodyn pciictration is inversely related lo wood density (MUNE- 111 & RAunfoNI, 2000). In llie two trials asscsscd, Pilodyn pencuation was much lnorc licritablc than growth (0.43 and 0.52). Tlicsc single sitc 1icrii;lbilitics were higlicr tlian those rcponcd hy M,\cDoN..~LI) et 01. (1997; 0.28 to 0.41) and ],articularly MUNERI & RAY- h i o ~ u (2000; 0. I3 to 0.27).

Pilodyn pcnc11-ation was liiglily gcnctically corrc- latcd (1, = 0.89 2 0.04) across the two sitcs assessed. This correlation was close lo the average (0.91) fl-o~n a study of five trials eslahlished in Tasrnaiiia which included miirc than 500 fa~nilies (MACDONI\I.II CI 01.

Tree form

Forking and tree forin arc not commonly iiicludcd in eucalypt breeding progralns as pri~inry sclcction trails when the ohjcctivc is pulpwood produciion (BORIIALIIO et nl. 1993; GREAVES ei iil. 1997: WE1 8.z BORRALI~IO 1999). However, such trails ?nay he imnportan: foi- solid \n,ood ohjcclivcs (R/w~l.ioNu 2000). I11 tlic Argentinian 11-ials, the early lei-king observed aitcr hail was poorly heritable (0.03 and 0.04) and the lrccs recovered aCta- a fe\vmonlhs. Howcvcr, this da~nagc did appear lo lin\,c impact on later age ionn (4 years) as the genetic c o w - laiion hctweeii the presence o i forks at 2 ycars and trcc form assessed at age 4 years was signifci~nt atid nega- tive at hotli sitcs (-0.40 and -0.79). Poor iorrnappearcd to a1 Icast partly arise irom tlie forking early i ~ i cslah- Iishmciit. However. the heritability of form on the Cour sitcs \\'as also low (0.07 to 0.11). VOLKER et id. ( I 990)

rcporicd higlicr hcrilahilit) (0.22 i 0.07) for fortii frorn :I trial assessed at ngc 8. 1-lo\b,cvcr ii i both studics. the licrit;ihiliiy of lrcc h n n was lo\\:a- Iliait es1itn;llcs lor gro\vtli. Dcspiic tlic low heritability 1(1r Forking and forni, tlie ge~ictic csprcssioii o l tlicsc mils was liiglily correlated across sites, at lcasl witliin suhfiiccs (Table 4: lorin 0.70: forking 0.77).

Vegetati~e phase change

In Iictei-ohlastic species such as Eirc(ii\?~iir.s g1ohiilir.s. tlic tinii~ig o l transition from juvenile to adult foliagc type is impoilant as llic leal' types differ iii numerous pliysio1ogic;il and aiiato~~lic chiimctcrisiics. 111 the case C I ~ E. glnhillio; many pcsls arc adapted to one or an- other iypc of foliagc (c.g. L,\wlcEuCE er iil. 2002; Srrixnt\uric 2001), and variation in the tiniing of the transition may affect~ieriorlnancc (JO1el)-iN eicii. 200~)). Tlic proportion of adult foliagc was the most heritable of all traits in the picscni study, a\,craging 0.60. High hcriiahility esti~natcs lor this ontogenclic trait wcrc siinilarly rcporicd in a base population trial of E. ginhii1ir.s in Tasiriania (l~cight io phase change it'= 0.67 i 0.13: JOICIIAN el (11. 1999) and lor oilier eucalypt species (WILTSIIIIII'. ef 01, 1998). Desl~ite tlie lower lict-icabilily ~scporlcd hy l l > i u z ~ er a/. (1994: Iieiglit 10

phase cliangc 11' = 0.30), the timing oitlic tsarisition to adult foliage s c c ~ t ~ s clearly to be under stroiig gciictic control in eucalypts. In addition, llic rclativc timing of this transition appears ~ I I he extremely stable across cnviroiime~ils. Tlic very high gcnctic col-rclatiotts hctwccn (0.92) and within (0.99) subraces lor tlic proportion of adult f<~liagc in the canopy. implies there is 110 cti \~iro~i~itc~it ititc~mctio~i afl'ecti~ig [lie expression of iliis o~itogcnctic twit (see also JOICI),\N ef 01. 2000).

Inter-trait correlations

Tiic gcnctic corrclalio~i hctwccn two trails will deter- ~iiinc how selection opci-ating on one trail will aflcct ccnctic variation i n anolhcr (i.c. corclaicd selection - - FAI.CONEIC & M..\cK,\Y 1996). A positi\,e gcnctic changc in ooc trait could arlcct otlicr tmils i n llic populaiioii in an adverse ot- Ihvoi~l-ahlcmaiincrdcpcnd- iiig upon ilic strciigth and diicciion of tlicir gcnctic c ! ~ r r c l ~ ~ t i ~ ~ ~ i s .

Willii~i mccs; gcnciic miriation i l i growtli was cllcctivcly itidcpc~idcnt o l gctictic variatioti iii lhc propo~-tio~t of adult fc~liagc i n lhc canopy, tlic lcvcl ~ I I I~rkiiig, Pilodyn pcnctratiiin and rclativc hark thickness c 6 Tlic otily exccptions i~ivolved specific gn)b\,tli cnl-relations witli vcgclativc pliasc chaiigc. Tlicrc was a lo\\:, hut significant positive corrclatioii of

early growlli uitli the proportion o l adiilt ioliagc at two sitcs. Tliis is indicative 01 tlic general trciid observed at the cnvit-o~imctilal and genetic level hy JOilD.4N er (11. 12000) for Ihst gmwtli to result iii more mpid 11-ansitioii to adiili kiliagc. This trcnd uas also evident in the genetic corrclntion ~ v i i l i lour-yeal-diameter rcportcd hy IP INZ~\ ef (11. (1994). hut tiot lot- tlicir carlicr age mca- surcrncnts. Tlicrc was no sigtiificatii associatioil be- t\vcen growth a~id tlic transitiun to adult foliage at the subracc level i n the present study. although soch a corrclatioii was rcporicd for a disease daiixtgcd site in T . , i r~n~i~iia , \b,hcrc early pliasc change resulted iti hcttcr gi-ou:lh (JORII~\N er 111. 2000).

Surviv;il was gcnctically correlated \viili gron.tli 11-ails at VOCA (Table 5). VOCA was tlic oiiiy trial wlicrc tlicre w-as a significant genetic basis lo variation i n surviv:il wiiliin subraces (Table I ) and, as with growth, survi\,al was aisocorrclalcd \villi tlie proportion of adult foliage. but not with ibr~n. The correlation estimates bctbr.ccn survival and growtli \veuc \nritliin the top range of tliosc reportcd previously (CII ,&~!I~EIIS et 01. 1996) and indicate inortaliiy was s i x dependent due lo tlie dcatli o l gcnctically siowcr groiviiig plants. However. tlicsc sigiiificant gcnelic coi-relatio~is (r,) \\.ere no1 cvidcnt at the subracc level (I-,; Table 5 ) .

Thcrc was a geiicrai tcndcncy fol- good lorm to be gcnctically correlatcil witli fast growlli at all sitcs (T;lhlc 6). Tliis association was not dctcctcd at tlic suhrace level (Tablc 5 ) hul a similar trend was also rcportcd by VOI.I<Elc ei (11. (1990) ii i E. gimhiiiii.~ wlierc goild lrcc foi-m was wciikly geitctic;111y ;~ssocialed with last growtli (1; lor height 0.43. diamc1c1- 0.07 and voluinc 0.13). Regardless o l whctlicr this trcnd is a scoring arliiact or a true biolc~gical tendciicy, ilic gcnctic correlation is clearly iti a direction iavoui-i~hlc lor bl-ccdcrs.

Tlicrc was a tendency Sol- n-ccs u~ith gctics lor tliickcr bark to also have denser wood. Tlic correlation hctwccn the relative bark thickness and Pilodyn pelic- tration was highly significant at hot11 sitcs assessed (1:

= -0.46 to 0 . 4 2 : Tablc 6). Tliis trcnd was also cvidcnt at the suhi-acc lcvcl (Tdbic 5) and has also bee11 oh- served in trials in Tasniania (DUTKOWSKI & P o l r s 1999), wlicrc it was suggcsicd to he a plciotropic rclatio~iship reflecting iroti~tlicjointorigiii olivood and hark in tlic camhiurn.

A knowlcdgc of Llie gcnctic rclationsliip hctwccn grc~wlli illid wood density is importatit lor brccdiiig programs. An adverse corrclaiioii, as ohser\,ed in some POiiis species (c.g. BURIION & LO\\' 1992). call mark- edly retard progress i n sclcctioti to\i-al-d a last growitig. liigli iicnsity idcotypc ~scquircd to optimisc pulp yield ~ C I - licctarc. Wiiliin siics in tlic present stiidy. Pilodyii pmciration was gc~ictically itidcpcndc~tt of diatnetcr

be made in iiiipi-ovcmcnt ol' this base l>opuiation l'or pull?wood pi-oducti~in in Argentina. A srroiig gciictic curl-clarion aci-oss agcs iind sites suggcsis that early sclcc!ion forgmwtl i and as inglc hrccdiiigpopuiatioii is

possihlc with lirtlc loss of genetic gaiii.

ACKNOWLEDGRIENTS

Tlie aulhors ilhank INTII- SOPORCEL (ltistitulo Nacional dc Tecnologia Afropecuaria. i\rpcniina - Socicdad Porriigucsa cic Palxi S. A.. I'i~itug31) for /;icililaiiiig tire data. Dr. Greg Jordan ior coniiiicnts and assis1:incc with analysis and Pcicr \'olker ibr allowinf us to tisc i~npublisiicd rcsiilts. 'Tii;ils and daia collection were li~ianciaily supporied by the Proyecio Forestal de Desnrr~~llii. INTI\, the Arfcntininn Government and Univci-siiy ol~lhsiii:rniaspiinsi~rcd a postgraduaie schol:ii- shiil ihr ihc sci,iiir :iuiiion

LITERATURE CITED

A I I ~ \ ~ ~ J O , J. i\.. LiD1OS. L., RnhlOS. A,. FERlIlilll,\. I . G & BoaR:\iiio, N. bI. G 1997: TIic RAIZ Eircoly~~ir~.~ globri/i,s hl-ecding program: i\ BLUP rolling-koni sirategy with ;I mixed cIon:il and sccdliiig dcploymcni scheme. l i i Proc, oiilrc IUFRO Conii.rcncc on Silvicult- ure and Improvement ol. Eucalypis. Salvador. IJaIiia, 13iasil. EMBRI\I'A, Coloinbo, lip. 371-376.

~~,\i.hlil.i.i, ti,, RI~SQUIN, F. & 'l'ilUJii.i.o, I . 2001: Esiiinaciiin dcpal-$iictrosgcnCiicos parlicaiaciciisiiulis dccrccimicn- to y prodiictiridad en I<~~ciily~~rirs ,g/oh~rli~s cn Limgiiay. I!,: Proc. of tlic IUFRO lotcrn:iiion:il Symposiuin on DevelopingilrcEitcalypt oiiiieFuiiil-c. INFOR, Valdivin, Cliile.

lJo~~n,\t.iio, N. M. G. 2001: Tlic purpose o i brecding ir hi-ccding ibr ;i jpuiposc. lii: Proc. o i thc 1U17R0 Inierna- tiiiriai Syinposiitni 0x1 Developing thc Eucalypi o i ihc Future. INFOR. Valdivia, Cliilc.

ROl?Ilh~.iiO, N. M. G, CO~l~lIll.1.. P. P. & KANO\\'SKI, P I. 1992a: Genetic control ofgroi\;th oSEiicaly/,irr.sg/obirirr.s in Portugal. 11. ESlicicncics of curly sclcciioii. Silvne Geiierkn 31: 70-77.

1301IRAL110. N, hi. G , Kl\NO\VSKI, P .I. & COTTEIIII.1.. P. P. l992h: Geiieiic control offrowilh ofC~icu/j/~iirsglobrii i is in Piil-iugal I . Geneiic and phenotypic parameters. Si1i.i~ Co>eiicii 31: 39-45.

~~Ol lKAI . I IO, N. M. G.. COTTlilllLL. P. P. & KI\NO\VSKI. P. J. 1993: B I . C C ~ ~ I I ~ oi>jecii~es iiir pulp produciion of Ilitco- l?/,iir.r glohi~lris ibndci diSl~roiii indusirial cilsi simctuics. C<,~wrdi~z .i,,ur,~r/ <jjFore.~i Re.~curc/~ 23: 648-656.

BORIII\I.HO. N. bl. ti.. J..\IIVIS. S. F & POTTS, 13. h,l. 1995: blibltivariliic i-csirictcd ininrimun~ Iikcliiiood cstiinaiiiin 01' netrciic narnmcten lor diameter ;ri 3 io 5 vcars in the

BOi~il..\i.tio, N. hl. G & POTTS. 13. M. 1996: Accoiinting Sui nativcstand characici-isiics inge~reticev3iua1ioi~s oSopeir pollinated pnigcny Srom liiicnl~~iiir. ,globi,lrrs basc populniinii. ~ \ ~ c ~ r ~ Fowsri 11: 53-61.

B a o o ~ i i i ~ , hi . I. M . 2000: A nciv classificntioti of ihc genus Eliciil~/~iri.s L'ller. (hgrtaceae). riitsri-nlinii Sysieiriniic Iliirn,:? 13: 79-1 48.

BURooN, R. D, 1977: Genetic correlation :is 3 concepi lor siudying genotype-eiliiroitmzni iotei:iction in iiiiest lice hrccdiiig. Si/~,iii, Geneiicn 26: 168-175.

Bulli>o?i. R. D. & Lo\v. C. B, 1992: Genetic survcy oSPiiiris niiliiiin, 5: wood properties: variaiioii. lheriiahilities. arid interrclaiiunsliips \villi oihcl- tl.:iits. ~\~eii~Zeii/iiiii/.l~i~l-~lNi

o/Foresr,?. Scieiicp 22: 228-245. Ruac;iiss, I . P . \l'li.l.i,\hlS, E. R., 13ii.I.,J. C.. il,\R\VOOD, C. E.

&OWEN. J. V, 1996: Tliccifecioiou~crossithf rntc oil ilic growtit oiselected lhniilies oil~iici~h.~~ii~.rgi-o,idis. Si/i,iic Gerieiica Is : 97-1 01.

Cti:~hliiiiRs,P G S. &Rollllni.lio.N. hl. G 19'17: lnlporiaticc of survisal in short-rotation iree breeding progmtiis. Criir<idiioi .loi,i.z~u/ O~JVITXIIS. Reseorcli 27: "i I -9 17.

CII,\hIRIIIS. P G S.. ROIIK..\LIIO. N. h'l. G & PO1'IS. B. hl. 1996: Grneiic analysis oisoi\,i\.ai i n i ~ ~ ~ c n / y ~ i n ~ s p I o h i ~ l ~ ~ . s ssp. globi~1ri.s. Sil\.ne Gciiericir 45: 107-1 12.

C l i h k ~ u ~ n s . !'. G. S., P o n s . B. M. & ' l ' i i ~ ~ ~ ~ i ~ o . P. 1997: The gmhclic ciinirol o i iloa,cring pi-eciiciiy in E,icn/y)~rrrs filohitlrrs ssp. glohirlics. Si/l,iie Geilpiicii 16: 207-211.

CORNlil.lUS, J , 1994 llciitahilitics and addiiivc fcnciic coeiiicienisoivariation in Soresiirees. Co!~ndL~~~.loit,iiiil ojFriie.si Rcseii,rlt 21: 372-37'1.

DUNGIY. I!. S.. POTTS, 13. 'I.. CAIINI;Glli, i\. J. &iii>iS, P K. 1997: iWjcos~~hn~,rlIn leaidisease: Genetic variaiion in damage io Errcilhptus ,iiie,is. E. ~lobliliis and their F , !hybrid. Ciuladiii,~ Jorrniol qf f i r -es i Resear/> 27: 750 -759.

DUTKO\VSKi, G. W. & POTTS, B. bl. 1999: GeofrnplIic:li paiierns oigeneiic variation in 1l~rcahpiic.r globi~/i,s ssp. g1ohi~lri.s olid a revised racial ciassiiication. Aiisii-<z/ii~i~ Joirnliil o/LIoin,iy 47: 237-263.

DUTKO\\'SKI, G. b'.. I'OTTS, B. bl. & 13011~~i.iio. N. h4. G. 1997: Rcviicd racial classificaiioii oSl.'~rcnI~pir~.r globrri- ~ri ssp. glohrili~s ;ind ihc impoit:iiicc of including race in aii:ilysis iiSpl-ogeny iiials. 1,i: Pi-iic. oilhi' 1Ul'liO Con- fcroicc on Sili.iciilturc and liiiprovcmcnl oSEucalypts. Salvador, Rahia. l3sasil. EMBRAPA, Cilliimho. pp. 322-329.

DUTKOWSKI. G W , GIL~IOUII . A. R. & ROIII<AI.HO. N. M. G. 2001 : hlodilication oi i lheaddi i iverc ia i io~is l~ ip~~hair ix SIX open p~iiinatcd iiials. iii: PIOL.. 01 the IUFRO lnicrnii- iioniil Symposiuin on Dcvclojiiiig ilic Eucalypt 01. ilrc Future. INFOR. Valdivia, Ciiilc.

DUTKO\\'SI<I, G. LV.. COST:\ E S~i\o\ . J.. G I I . ~ I O U R . A. R. Br Lo1,i-z, G. !\. 2002: Sp:itial analysis metlioiir ior hi-cst gcneiic iri:iIs. Cniziidin~i .iiirii?>iil iijFiireir iles~~nil.li, (iit

"

Ei~cnlyl~iirs globi,/irs base populaiion in husiralia. i i i : - ~ ~ r ~ ~ s . s ) . ~ u c ~ i ~ y p i ~ ~ a n t m i o n s ~ ~ m l , r o i i , 1 g ~ : ~ ~ r e ~ ~ i e ~ ~ a n ~ ~ u a ~ ~ i y , t~l.l~Rli!Gli, K., DAVIDSON. J., HR\VOlOI j , C. & \'\N WYK, G

CRCTt1I:-IUFRO Cunierence. (Ed. B h l . I'oiis, N h l G . 1993: Eocalypt Domcsiicntion and Rrccding. Clnrcndoii

Boi-i-alho. J.B. lieid, R.N. Cromer, \VN Tihhits and C,\. Press. Oxfiird, 288 pp.

ply i,,o,, d , ) Hob:iii, ~ ~ , ~ , , ~ , , i ~ , CRC ~empci-a tc EIIICSSON, T. 1997: Enhanced llcritabiliiics and licsi Linear

Hiil-durooii Foi-csiry. pp. 230-232. Unbiased Predictors Through Ajipi-opi-iatc Bliicking o i I'rogcny Trials. Coiitidiiiii Jorini/i/ q/' i;o,r.si Rp.seiiir11

27: 2097-2101. FAI.CONI:R. D. S. & M,\cK.~Y, T F. C. 1996: Ii~tiodu~!ioii lo

Quantitativc Gcnctics. Addison Wcslcy Longinan Liin- ited, Edinburgli Gatc. Ilarlow. England, 464 pp.

Fu, Y H., Y,\NCIIUK, A. D. & NA~IKOONG. G. 1999: Iticoin- pletc block dcsigiis for gcnctic testing: soinc piactical considerations. Cannilion Jortuinl of' Forrsr Researcli 29: 1871-1878.

Gtt,z$oun. A. R., THO~II'SON, R., CUI.LIS. R. R. & \VII.HA~I. S. J. 2001: ASREML Rcfcrcnce hlaiiual. Technical Report NSW Agriculiure.

GREAVI?~, R . L., BORR~LIIO, N. M. G & RAYMOND, C. 4. 1997: Breeding objeclivc for pl:intatii>n cucalypls groiln for production of kraft pulp. Forest Scieitcr 43: 465 1 7 5 .

GRII:I:IN, A. R. XL COTTERILL, P P 1988: Genetic variation in growth ofoutcrossed, selfed and opcn-pollinated progc- nics of Et,cnl~piirs veg~~rnis and some iinplications for hrccdinf strategy. Silvne Ceiiciicu 37: 12L131.

GUTI~RIIEI, B..MOLINA, ~'I.,BALoccIII,C. &DROPI 'EL~~ANN, E 2001: Desarrollo de una iaza local dc E~<calypt~ls glohirlirs tolcranlcal Trio. I,!: Proc. oftlic IUFRO Intcriia- tional Symposium on Developing tlie Eucalypt of the Future. INFOK, Valdivia, Chile.

MI\IIU~KU. J. L., GRI~:FIN, A. R. & ESI'EIO, J. 1999: hlass controlled pollination o fE~~cn ly /~ i i r sg lob i i l i ~s : apractical reality. Cnitnilion . l o i~nml of l;orcsi Rescnicli 29: 1457 -1463.

I-IARDNER, C. M. & P o n s , B. M. 3995: Inbreeding depres- sion and changes in variation aner selling in Ettcnhpitrr glubulrrs ssp globrrlirs. Sil~xze Geneiicn 43: 46-54.

I~I,\RDNLR, C. M., POTTS, B. M. & GORE, i'. L. 1998: The relationship bctu,ccn cross success and spatial proxiinity ofE~,calypirrs g luba l i~r ssp, glvbirlirs parents. Evoletion 52: 614-618.

HARDNER.C. M., Vr\lLI.ANCOURT, R. E. &POTTS,B. M, 1996: Stand dcnsity influences outcrossing rate and growth or opcn-pollinated Families of E~icnl\prus fi lobii l~is. Silvnr

~ ~ ~~ ~

Gerlericn 45: 226-228. HODCE. G R.. VOLKER. P W.. POTTS. B. M. & O\VliN. J. V

1996: A comparison of genetic information from open- pollinated and coiitrol-pollinated progeny tcsts in two eucalypt species. Tiieoi-eticnl mid Applied Gr,ierics 92: 53-63.

Hout.~, D. 1992: Coinparing evoivilhility and \,ariability of quantitr~tivc traits. Genericr 130: 195-204.

IPINZA. R. H. , GAIICL, X.; AI'IOIAZA, L. A,. ~IOI. INA, hl. P , CHUNG. P. & PARRA, P. 1994: Variaciiin juvcnil dc uii cnsayo de procedcncias y fainilias de E~rcn/ypti,r xlo- hrtlrrs subsp. globirlirr Labill, en la s6ptinia regiiin, Chile. 1~~:oIog iu 8 : 259-270.

JORD:\lN. G I . . POTTS. 13, bl. & WtITSHIIIE. R. 1999: S t r o n ~ . - independent quanti!ati.i,c genetic control of vegetative phasc changc and first floi+,ering in Eucnlyl~tics gloh~,liis ssp. globrdus. Plcrudiry 83: 179-187.

JORDAN, G. J., POTTS, B. M., C H A L ~ ~ E ~ < S , P. & WILTSHIRE, R. J. E. 2000: Quantitativc gcnetic cvidcncc that thc timing of vcgcmtivc pllasc change in Eircalypiirs g lob~~ l i r s ssp. globrrlus is an adaptive trait. Air~.lra/ia,i Jo i rn~a l OJ

Roiany 48: 561-567. 1ORI)AN.G J.,POTTS, B. hl. &CI.I\RKII, A. R. 2001: Susccpli-

bility of Ei rco l y~~ t i~s globrrli~s ssp. glvhic1ri.s to sawfly [ P e r ~ n nfliniiiiis ssp. i,isu/aris) attack and iis potcniial impact on plantation productivity Fo,-est l~cnlo,yy orid ~t/~oingeiizp,tr 160: 189-199.

KELtu. C. hl. 1997: N:itur;~l \-ari;ilion and gciictic control of relative bark thickness in E,rcnhptirs g l o b i ~ l i ~ s ssp. glohi~li is. Unpuhlishcd BSlHonoursTlicsis. University of Tasmania.

KlRKl?.YrRlCK, J. B. 1975: Geographical \'ari:ition in i:!,ciiIy,>- r~ts glohrrlirs. Forestry ;~nd Timber Hurcau. Bulletin 1 7 Australian Gor.crnn?ent Publishing Service. Canberra. 64 PP.

L:\Y\'KtENCk R., POTTS, B. M. & WIIITHA~I, T G. 2002: Relative imporiancc of plant oniogcny, lhosl gcnciic variation, and leafage for aconlnion lherbi\~orc. IEcolngy, (L, press).

Loi'tiz. G. A , AI'IOII\%A, L. A,, P o n s , H. M . , DUTKO\\'SKI. G W., GEI.ID, P & RODRIGUEZ TRVIRSO, 1. M. 2 0 0 1 ~ : Genctic pammetcrs hi growtli and pilodyn from Eilca- Iyj~r8is gloh~il irs in Argentina. I,,: Pi-oc. of llic IUFRO l~~tcrnatiunal Syrnposiurnon Dcvclopin~ thcEucalypt of tlie Future. (Ed. S. Barros). INFOR. Vaidivia, Chile.

Lot ,~z , G. A., POTTS, B. M., DGTKOWSKI. C \V. & RODRIGUEZ TRI\VI?RSO, I . hl. 200lb: Q~iant i ta t i~c gclletics of Erri:u- 1yptic.s glohiclits: Afiiliities of land race and naiivc stand localilies. Silvae Ge,icticn 50: 256252 .

LOPEL, G A,, P o n s . B. M., RODt<tGUmTtI.4\~ERS0, J. M. & GELID, P. 2001c: The perfc~rmancc oCEiccahptits ,nnide- rlii pIo~,enances in Argcnliila. /,I: Prilc, of tlic IUFRO liiternationai Sympoiiuin on Dcveioping the Eoc~lypl of thc Future. INFOR. Valdivia, Chile.

LOPEZ, G A,, POTTS, 13. M. &TILYARU, P A. 2000: F, hybrid inviability in Eucnhpr~is: tlie case of E. nviitii x E. globules. Her-ediv 85: 242-250.

MACDONALD, A. C., BOI<RAI.HO, N. M. G & POTTS. H. M. 1997: Gcnclic variation ibr growth and wood density in 1: l~nlypius gluhslus ssp. glvbirlirs in Tasmania (Austra- lia). Sil,,oc Ccizeiicn 46: 216-241

M U N ~ ~ R I , A. & R , \ ~ n l o ~ m , C . A. 2000: Genctic paraiiictcrs and gcnotypc-hy-cnvironnm intcractions for basic densit): pilodyn pcnctraiion and stem diammcr in Etrcal?/~tir.s glohulirs. Foresl Generics 7: 3 17-328.

POTTS. B. M. & JORDAN, G I . l99ia: Gcnctic variation in the juvcoilc leaf rnorpliology of !~~,<:ulypi~rs gioh~r1ii.r Libill. ssp. glohicliu. Forest Ge,ierics 1: 81-95.

POTTS, 8 . h1. & JORDAN, G I . 1994b: The spatial patlcl-n and scale of variation in E~rcnljpi irs glubirlus ssp. globr~lirs: variation in seedling abnormalities and early growth. Ait.sir-cilian Jurrrrlol of'l3otnny 4 2 4 7 1 4 9 2 .

POTTS, B. M., VOLKI;R, P. W.. HOUCE. G. R., BORRAI.IIO, N. M. G , HhRDNER, C. H. &O\VEN, J. V 1995: Gcnclic s to the exploitation of basc populations of E:'ircn/yi,nrs glohirlitsssp. globuli~s. In: Eucalypt p1;intations: In~prov- ing Fibre Yicld and Quality. Proc. CRCT I-IF -1UFRO Cont , Ilobart, 19-24 Feb. (Ed. R .M. Ports, N.M.G Borralho, J.B. Reid,R.N. Cronier, \VN. Tibbits andC.A. Raymond.) Hobart, Tasmania. CRC for Temperate Hardib'ood Forestry pp. 217-221

P o n s , H. M. & WIITSHIRL:. R, J. E. 1997: Eucalypt genetic and gc~~ccology. /,I Eucalypt Ecology: Individuals to Ecosystems. Ed. J. Williams arid 1. Woinarski.) pp.

56-91 Cmiihridge University Press. Cambridge. l < : t ~ h t o ~ ~ > , C., \ 2000: l 'iee biccding issucs lor solid aroiirl

prodiiciioii. lii: l'roc. uftlic IUFRO inlem;iiii~n~rl Conici- eiicc on Thc 1:uture of Eucalypis Sor \\'ooii Pi-oducis. L.iiinccston. '1'asin:inia. .4ustialia, pp. 265-270,

RE%EN!)I', G D. S . P & r > i RI~SENIII:, hl. D. V. 2000: Domi- nance ellkcis in /~ticn/?l~irisgi-ii,idis, Ericiihpii~r i~i?~p/ryl- la axid lhybiids. lir: !'roc. oStlic Symposium on llyhi-id 13recding and Ccnctics. (Ed. I I . S. D. Dungcy, hl. J. Dieierr and D. G. Nikies.) Brisbane. Dcpartmeiit o i Piiiixti-y lnduiti-ics pii. 93-i00.

S,\NlltiEii:\, R. &G!iil:riN. A. R. 2001: F,\hI,\SA Fiber Yiclil Improvcmciii Prograninie (EY.I.I'.): 10 yclii-s cxpei-iencc biecding l l i r c i i l~~~ i i i s ,g/ohi~lrr.s. I,,: Pioc. 111' iiic IUFRO liiici-n;itioo:il Syiiiposiiim on Dcvclopit i~ i1ic Eucalypt o i tiie Fuiiirc. IlYFOR. \'aldivia. Cliilc.

Sii,\iii.ii,S.R. I971 : 1.iticarhlodcls. Joliti \\'iley & Soils. Ncm h r k . 532 jip.

SK~IIIO. S.,V~it.t.:\NCoti%T, R. B. &POTTS.B. hl. 1998: Fine- sc:ilc gciictic su-i~ciitre of l<,<cnlypilr. ,glohii/rr.s ssii. fi/iihiili<s lbrcst rcvc:iIcd by RAPDs. Air.sii-ctliiiii Joiir,>ii/ i>/'!loiiiii? 46: 583-591.

Soilih, F.. H,\s~~Rco. F.,Tov,\i.. G, SII.IO, L., RODRIGLIE%, 11. C. Sr Tol<O, hl . 1997: Bayesian eslirnation 01. gcnciic para#nctwmnd Iprovenancc cifccis hi height and diame- tci- of I.'itcti/?/~iiis glohi,/ir.s in Spain, lii; Pmc, o i the IUFRO Cniiicrcncc oil Silvicullurc and liiiproveineiil o r Eucalypis. S:iI\,ador, B;ihia, Rrasil. EMBR/\P,\. Co- looibo. pp. 95-100.

Sonl,\, F & B o ~ i t , ~ ~ i l o , N. M . G 1998: Tlic gciiciics o i I-csistr~ncc lo l'liorriciiizilio scii~ipiuicinia attack in Euc.a- I?~!us sIuIII~II~s iii S1)iiiri. Si1i:nr Getieficn 46: 365-3159,

S'riiiNu~\titili. h'l. J. 2002: Os,iposiliiiii prcici-cncc and neonate pcrloriiiaiicc i l l ,Vl~ie.sriri?/,e/o pi.iviiiii in rclalioii lo lictcrupliylly in I.'ircii/?~niw d,rrtiiii and E. glohrili~s. Agrici~/ir,riil iC h i - t s i E>:iiioi,iology, (if, 11i.i.s~).

'~III!IITS.\V. N.. 1300hlShlr\, 0.13. &J!\R\'IS.S, 1997: Distrihu- Lion, b io io~y , gciieiics, and ilnproveineni programs lor Iliicnhpfiis glohi~lirr and I:. iiiietis around the world. Iri Proc. of ilie 21th Biconial Snilthc~n Ticc ltiijiruictiietii Caiiicicncc, June 9-12 1997. !<</iii,i/ ii). T \Vhite, D. lkluhci and G Pou,cll.) pp. 81-95, Soi~thcrn Tree Im- pi-ovcrncni Coinmiitcc, Oi-l;indo, Florida.

'Tijiu), \I. A,. SI I IO . L., Roon icu i z , h'l. C.. Sea!!\, I;. Sr G 1998: Genetic aii;i!ysis o i suni\'al lo droiiplit

iii Eircii/?l,iri~ ~1obirIii.s in Spain. lii: Proc. of the 6"' \\Torid Congress oii Genetics i\pplicd to Livestock Prodiiciion. jp)' 499-502.

V l \ i i . i . h ~ c ~ ~ ~ ~ i ~ ~ . R. E., POTTS. 8 . h.. \\'ATSON. hl.. VOI.KI~II, P. \V., llC)l~Gli, G. R.. Rl!ii~. J. R. & \Vl:SS, ,\. K. 1995: Dcicctio!~ and prediction of hcterosis in E,~CL~/!/JI~<.T ,gl<,/>i~/~,,r, Fore.?f G < W I ~ I ~ C , T 2: 11-1 9.

\ ' ~ ~ t . K l ~ t ~ . P \?', D!?,\N, C. A,. TiiiiiiTS, W. N. & ~<:\VliN\\.0011, I . C. 1990: Genetic paialnetcis ;inii gains c r ~ ~ e c t c d li-om sclcctioti in Ericnly,,itu ,plolii,/irs i i i 'C:~sin:inia. Silviie

describe age trends in lhcl-iiahility in sl~ort rotatioii lice specier. lii: Prou. oTihe QFRI-IUFRO Coiiii'reiicc Ticc Improi-cmcnt for Sustainable Tropic:il l~orcstry. (Ed. %I. J. Dieters, ,\. C. hlailicsuii. D. G Niklcs. C. E, l l;il-u;ood and S . hl. Walkel:) Caloundrn. Queenslaiid, Australia. 1996. pp. 178-181.

\VEl. X. & BOiiit,\t.iio. N, hl. G, 1998: Usc ofindiiridual trce iniacd miidcis to ;iccounl ior iiioriality and selective thiniiiiigwhen esiimating basepopi~iatioogciiuticp:ir~im- elel-s. Fo~oi-evf Science 44: 216-253.

\Viii, X. & I3oRR:\!-!!a, N. hl . G 1999: Olhjectivcs and selec- tion criteria lorpulp production o i l i ~ r c i i ~ , i i r s i i i i i ~ ~ l i ~ l l i i

plaiitatioiis in So~itlr East Cliin;~. Foi-eii Ge,ieiici 6 : IS1 - I 90.

\VIIITIi. T. 1996: Genetic parameter crtiinatcs aiid hiecdiiip value picdiciiiins: issues aiid irnplicaiions i i i trcc iin- pl-i~vcmcnt piogianis. I,,: Proc. o i the QFRI-IVFRO Co!iicmnce Tree liiiprovemenl for Sustain:ibIc 'I'ropical Forestry. (Ed. hl. J. Dictci-s, t2. C. \lailicson, D. G Niklcs, C. E. tlnrwood arid S, hl. Walker.) Ceioundra, Quccnslaiid. Australia. 1996, pp. 110-1 17.

\Vii.l.i:\hts, D. R.. P o n s . 13, hl. & Bi.,\CK, I'G. 1999: TcsLing single visit polliiiaiioii procedui-cs for E i~cnhp i ru~ lo l i i r - /r(r and I<. ,~ire,?s, A~rsri.ii/ia,i i;ii,e,sii-y 62: 346-352.

Wll.Ts1llnE.R. J . E..l'o~Ts, B. PI. &RiiIo, J. B. 1998: Gcnctic control o f repioduciive znd vcgc~it ivc pliasc change in ihc E,,cn(vp:lrs risiliiiiii-E, loiirir-iiinir. coinpiex. Aii.sliri- linrr Juirninl of Bnrnriy 46: 45-63.

WOOL,\SSON, R., 'r\'irtTrhini.r, P ti. & C A ~ ~ I ~ R O N . J. N. 1991: Gciiciic :tn:ilyscs of b ~ i r ycnr diameter measures i'or l~,rc~ilip:ris g/oh~i/irs growiiig across lour sites iii Vicio- ria. Aoslralia, 111: Pruc, of the IUFRO symposiuin or1 Intcnsivc Forestry: The Role of Eucalypis. (Ed. A.P.G. Sclivnaii.) Durbliii, Soiiih Aliica. Soutiici-1, Ali-ican lnstiiuie oiForcstl-y, Prctol-ia. pp. 371-3x1

Z,\NG. I)., WANG. H. & YOU, Y. 1995: PcrlOrma~~cc and seleciion o i a 4-year fircri/y/,i~r,s g/iibii/ir,s sccdliiig sccd orchard in Yuiiiiaii, China, I,,; 'Eucalypt plnntaiions: Improving Fibrc Yicld rind Quality' Pmc. CRCTIIF - IUI'KO Cooi:; Hobart, 19-2d Fch. (Ed. R. \ . Potts, N. hl . Ti Borialiio. 1 . 13. Rcid. R. K. Crotiicr. \V. N. TihRits :ind C. A. Raymond.) llobart. Tasmania. CRC for Tcm- peraie I lardii,ood Forestry pp. 226-229.

ZiiNG. Z.-H. 1987: Long-term col-rc1:itcd rcslioiisc, interpupil- Iiiiioii co~ariaiiciii. and ir i ler~~~ccii ' ic allometry. Evolirtio,~ 12: 363-374.