Journal of TROPICAL V Ol

JOURNA L

TROPICA L

FOREST

SCIENCE Journal of TROPICAL V ol. 31 No. 2 April 2019 FOREST SCIENCE

Volume 31 Number 2 April 2019 FRIM

Ministry of Water, Land and Natural Resources JOURNAL OF TROPICAL FOREST SCIENCE JOURNALGUIDE OF TROPICAL TO AUTHORS FOREST SCIENCE GUIDE TO AUTHORS JOURNAL OF The Journal of Tropical Forest ScienceGUIDE (JTFS) TO is AUTHORSText should have the following headings: JOURNAL OF an internTheation Journalal revie wofed T jropicalournal cForestoncer nScienceing the s(JTFS)cience ,is introducTteioxnt ,s hmoauteldri ahlsa vaen dth me eftohlloodwsi, nrge shueltasd aingds : JOURNAL OF The Journal of Tropical Forest Science (JTFS) is Text should have the following headings: teacnh ninotleorgnya atinodn adle rveevlioepwmede njot uorfn traol pcoicnacl efornreinstgs tahned s fcoierenscte , diiscntursosidounc. tPiroint, hmeaadteinrgias lisn abnodld mtypeeth aonds s,u rbe-hsuealtdsi nagns d TROPICAL FOREST SCIENCE an international reviewed journal concerning the science, introduction, materials and methods, results and prtoecdhunctosl.o Tghy ea njodu drneavle wloeplcmoemnet so afr ttrioclpeisc rael pfoorretisntsg a onrdig fionrael st ind italics.iscussio Authorsn. Print hshouldeading uses in bSIo lunitsd typ eand and appropriate sub-heading s technology and development of tropical forests and forest discussion. Print headings in bold type and sub-headings TROPICAL FOREST SCIENCE fupnrdodamucetsn. tTahl eo jro uaprnpal iweedl croemseeas racrhti colens rtrepoopritcinalg foorrigeisnta l symbols.in italics. Keep Authors footnotes should to theuse minimum.SI units and appropriate TROPICAL FOREST SCIENCE products. The journal welcomes articles reporting original in italics. Authors should use SI units and appropriate bifundameology, econltoagl yo, rc haepmpliisetrdy , rmesaenaargcehm oenn ttrop, silviiccaull tfuorree,s t symbols.Ref eKeeprence footnotess should btoe citthee dminimum. in the text as follows: An international and peer-reviewed journal published quarterly by the Forest Research Institute fundamental or applied research on tropical forest symbols. Keep footnotes to the minimum. cobnioselorgvya,t ieocno,l ougtyi,l ichezatimonis tarny,d m parnoadguecmt ednetv, esliolvpicmueltnutr.e , ‘Owens R(2ef00er2e)nce ....s’ shoro ‘u..l.d. b(eT ecitsheodm ine th&e P textetty a 20s fol00lo)’w. s: MalaysiaAn international and peer-reviewed journal published quarterly by the Forest Research Institute biology, ecology, chemistry, management, silviculture, References should be cited in the text as follows: An international and peer-reviewed journal published quarterly by the Forest Research Institute conservation, utilization and product development. Se‘Ontwenecness (2sh0o0u2ld) p..r.e.’f eorra b‘l.y.. .n o(Tt estsahrot wmiteh & re fPeertetny c20es.0 A0l)l ’. Malaysia conservation, utilization and product development. ‘Owens (2002) ....’ or ‘.... (Teshome & Petty 2000)’. Malaysia reSfenretennceces sc istheodu ilnd tphree fterxat bmlyu nsto bt est alirst ewdi tihn raelfpehreanbecetisc.a Al ll Sentences should preferably not start with references. All For correspondence and editorial enquiries relating to submission of papers and to journal orrdeeferr oenn cseps acritaetde pina gthese. tFeoxrt tmheu slits bt oe fl irsetefedr ein caelps,h aadbheetricea l references cited in the text must be listed in alphabetical subscription,For correspondence please e-mail and editorialor write to:enquiries relating to submission of papers and to journal and quality. clorsdeleyr t on th se pfoarllaotwe ipnagg eexs.a Fmoprl tehs ef olirs tj oouf rrneafel,r penroccees,e addinhgesr e For correspondence and editorial enquiries relating to submission of papers and to journal order on separate pages. For the list of references, adhere subscription, please e-mail or write to: and quality. andclo sbookely to citation: the following examples for journal, proceedings subscription, please e-mail or write to: and quality. closely to the following examples for journal, proceedings KA Sarifah Submission of manuscripts and book citation: Submission of manuscripts and book R citation:A, SA, MC & JD. ForestKA Sarifah Research Institute Malaysia (FRIM) ManuscrSubmissionipts should b ofe amanuscriptsddressed to the Editor of RA, SA, MC & JD. KA Sarifah RA, SA, MC & JD. 52109Forest Kepong Research Institute Malaysia (FRIM) JTFS, ForestMan Researchuscripts s hInstituteould be aMalaysiaddresse d(FRIM), to the E 52109ditor o f photosynthesis of seedlings of two rainforest Forest Research Institute Malaysia (FRIM) Manuscripts should be addressed to the Editor of 52109 Kepong Kepong,JTFS, Forest Selangor Research Darul Institute Ehsan, MalaysiaMalaysia. (FRIM), Submissions 52109 dipphteortoocsayrnpt hsepseicsi eosf. sFeuendctlionngasl oPfla tnwt oB iroaloingyf o2r9e:s t Selangor52109 Kepong Darul Ehsan JTFS, Forest Research Institute Malaysia (FRIM), 52109 photosynthesis of seedlings of two rainforest musKepong,t be ac cSelangorompanie dDarul by a cEhsan,overin gMalaysia. letter th aSubmissionst includes: 505–515.dipterocarp species. Functional Plant Biology 29: MalaysiaSelangor Darul Ehsan Kepong, Selangor Darul Ehsan, Malaysia. Submissions dipterocarp species. Functional Plant Biology 29: Selangor Darul Ehsan mus•t be declarationaccompanie thatd by thea co manuscriptvering lette rhas tha nott in cbeenlude s: 505–515. R, SS & E. 1996. Putative Malaysia must be accompanied by a covering letter that includes: 505–515. Malaysia • pubdeclarationlished (ev thaten in the pr omanuscriptceedings o rhas as wonotrk beening ectom Ry,c orrh iSzaSl &fu ngi of Paso hE F. o1r9e9st6 .R Pesuetravteiv, e Editors: KA Sarifah ([email protected]) & S Vimala ([email protected]) • declaration that the manuscript has not been R, SS & E. 1996. Putative paper)orpublishe offeredd (even for in prpublicationoceedings elsewhere or as work ining Necgtroi mSeymcobrirlhaniz,a Ml faulanygsia o. fP pP a9s6o–h1 0F4o irne sLt eRee SsSe revte , Editors: KA Sarifah ([email protected]) & S Vimala ([email protected]) published (even in proceedings or as working ectomycorrhizal fungi of Pasoh Forest Reserve, ConsultingEditors: KA editor Sarifah: FSP ([email protected] Ng ([email protected])v.my) & S Vimala ([email protected]) supaper)orbstantial lofferedy the s aforme publication or abbrevi aelsewhereted form ,in alN. (eegdrsi) S Ceomnbseirlvaant,i oMna, lMayasniaa.g Pempe 9n6t –a1n0d4 D ienv eLloepem SeSn te t paper)or offered for publication elsewhere in Negri Sembilan, Malaysia. Pp 96–104 in Lee SS et Consulting editor: FSP Ng ([email protected]) eistuhbesrt ainnt ipalrliyn tth oer s aemleec toror niabcbarlelyv i(aete.gd. ftohrem , ofa Fl.o r(eesdt Rs)e sCoounrsceersv Partoiocnee,d Minagns aogf ethmee nMat alnady sDiaev–eUlonpitmeedn t Consulting editor: FSP Ng ([email protected]) substantially the same or abbreviated form, al. (eds) Conservation, Management and Development internet)either in print or electronically (e.g. the Kionf gFdoormes tP Rroegsoraumrcemse P Wroocrekesdhionpg. s2 o1f– th24e MaOctloaybseiar –1U9n96it,e d EDITORIAL ADVISORY BOARD either in print or electronically (e.g. the of Forest Resources Proceedings of the Malaysia–United • internet) KualaKing dLumpuom Progr.amme Workshop. 21–24 October 1996, MEDITORIAL Abdul Latif ADVISORY BOARD internet) Kingdom Programme Workshop. 21–24 October 1996, MEDITORIAL Abdul Latif ADVISORY BOARD • of the manuscript especially its contribution to KualaSS. 19 Lumpu91. Effre.cts of multiple environmental ForestM Abdul Research Latif Institute Malaysia, 52109 Kepong, Selangor Darul Ehsan, Malaysia. [email protected] • Kuala Lumpur. FoMr estAbdul Resea Latifrch Institute Malaysia, 52109 Kepong, Selangor Darul Ehsan, Malaysia. [email protected] of the manuscript especially its contribution to stre SssSe.s 1o9n9 n1u. tEffrieenct tasv aoifl ambiulilttyi panled eunsevi. rPopn 6m8e–n88ta l PSForest Ashton Research Institute Malaysia, 52109 Kepong, Selangor Darul Ehsan, Malaysia. [email protected] of the manuscript especially its contribution to SS. 1991. Effects of multiple environmental PSForest Ashton Research Institute Malaysia, 52109 Kepong, Selangor Darul Ehsan, Malaysia. [email protected] • names, addresses, telephone and fax numbers, ins tMresosoesn oeny nHuAtr,i eWnitn anvawilea bWiliEty &an Pde ulsl eE. PJ p( e6d8–s)8 8 HaPSrvard Ashton Herbaria, 22 Divinity Avenue, Harvard University, Cambridge, Massachusetts 02138, USA. stresses on nutrient availability and use. Pp 68–88 PS Ashton • anndam ee-ms, aaidl dardedssreess, stesle opfh othnee acnodrr feaspox nunmdibnegr s, Reisnp oMnsoe ofn Pelay nHt tAo M, Wulitinpnlew Setr esWseEs. A&c aPdeelml iEc JP r(eesds,s ) [email protected],Ha Windmillrvard Herbaria, Road,rvard.edu Chiswick,22 Divinity W4 A venue,1RN, London, Harvard UK.Universit [email protected], Cambridge, Massachusetts 02138, USA. • names, addresses, telephone and fax numbers, in Mooney HA, Winnwe WE & Pell EJ (eds) Harvard Herbaria, 22 Divinity Avenue, Harvard University, Cambridge, Massachusetts 02138, USA. authorand e and-ma icontributingl addresses o author(s).f the corresponding SanRe sDiego.ponse of Plant to Multiple Stresses. Academic Press, P [email protected] rvard.edu and e-mail addresses of the corresponding Response of Plant to Multiple Stresses. Academic Press, [email protected] author and contributing author(s). San Diego. NNaturalisPat uBaasralis BBiodiversityiodiversity CCenteenterr, ,H Herbariumerbarium D Division,ivision, P .PO.O.. B Boxox 9 59517,17, 2 3230000 R ARA L eLeiden,iden, T hThee N eNetherlands.therlands. Pi [email protected]@naturalis.nl Manauthoruscript sand sho ucontributingld be submit tauthor(s).ed electronically via TSanable Diego.s, figures and photographs should be P Baas Th e total number of words (text + + tables PRGN Bodeker aCraneturalis Biodiversity Center, Herbarium Division, P.O. Box 9517, 2300 RA Leiden, The Netherlands. [email protected] the OpenM Jaonuursncarli pStyss sthemousl dp abgee s autb hmttiptt:e/d/ ewlwecwt.rmonyjiucarlnlya lv.ia compiledT oanb lseesp,a rfaitgeu srheese tas nadnd p nhuomtobgereadp hcos nsshecouutlivde lby e Naturalis Biodiversity Center, Herbarium Division, P.O. Box 9517, 2300 RA Leiden, The Netherlands. [email protected] + equaMtiaonnus)s cisr ip6t,0s 0sh0o ourld l ebses snuobtm ciottuendt ienlegc throen aicbasltlrya vcita. Tables, figures and photographs should be CarlGPRreen WCrane T Knoblochempleton JrCollege, Dean, UniversitySchool of Forestof Oxforyr d,& OxfoEnvironmentalrd, OX2 6HG, Studies, UK. geYalerr [email protected], 195 Prospectrvard.edu Street, New Haven, mtyh/eo Ojs/piendn eJoxu.prhnpa/l SJTysFtSe.m Ms apnaugsec raitp htst tmp:u/s/t wbew wsu.mbmyjiuttrenda l. ancdom repfielerdre odn tsoe pinar athtee s hteexett.s Danrdaw n ugmrabpehrse du csionngs eEcxuctievle. ly PR Crane tMhea nOupscernip Jtos,u srhnoaull dS ybset esumbsm pitatgeed ealte hcttrtopn:/ic/awllyw vwia.m thyeju Ornpaeln. compiled on separate sheets and numbered consecutively CTPRCarl 06511Crane W Knobloch USA. pete [email protected] Dean, School of Forestry & Environmental Studies, Yale University, 195 Prospect Street, New Haven, inm My/icorjso/siondft eWx.oprhdp. /PJTDFFS . Manuscripts must be submitted Maankde rseuferrer efdig tuor eins tahree tcelxeta. rD. rTawh eg rpaapghes ufosirnmg aEtx ocef l. Carl W Knobloch Jr Dean, School of Forestry & Environmental Studies, Yale University, 195 Prospect Street, New Haven, mJouy/ronjals/ iSndysetexm.psh page/JT FaSt . hMttapnu://scwrwipwts. mmyujustr bneal. su mbmy/iottjesd/ and referred to in the text. Draw graphs using Excel. DPOakCT Dykstra Spring06511 USA.Garden pete Foundation,[email protected] 1776 Loughborough Lane, Upperville, VA 20184, USA. [email protected] 12in-p Moiinctr ofsoonft sWizoer,d d. oPuDbFl e -space text and begin the thMe ajkoeu rsnuarle s hfigouulrde sb ea rcoe nclseidaer.r eTdh ien pdaegseig nfoinrmg atht eo f CT 06511 USA. [email protected] iinn dMexi.pcrhops/oJfTt FWSo. Mrda.n PuDscFri p ts must be submitted in Microsoft Make sure figures are clear. The page format of BlueDPDP Dykstra OxDykstra Forestry, 1602 Windstar Court, Paso Robles, CA 93446, USA. [email protected] m1a2nu-psocirnipt tf wointht stitzle, p daogue.b Alell-s ppaagcees mutexstt abned nu bmegbienr etdh e drthawe injogusr. nKael esph ofouoldtn boet ecos tnos itdheer emdi ninim duemsig. nDinetga itlhs e DP Dykstra 12-point font size, double-space text and begin the the journal should be considered in designing the SABlue Ha rOxris Forestry, 1602 Windstar Court, Paso Robles, CA 93446, USA. [email protected] conmWasoenurcdu.s ctPirDviepFlty .w Litihn a etri etnl enu opmta abgcecr.e sAp oltlan bp laeg.a ecUsh smu ep 1a2sg-te pb ooeif nnu tt efomxntb ta esrriezee d, gidveranw iinn gths.e Kteexetp s hfooutldn ontoets btoe rtehpee amteindi minu tmhe. tDableteasil s 8762Blue S OxW Fi Forestrviewr Place,y, 1602 Beave Windstarrton, Court, Oregon Paso 97007, Robles, USA. C Adennisdykstra@blueoxfo 93446, USA. [email protected] ry.com manuscript with title page. All pages must be numbered drawings. Keep footnotes to the minimum. Details DSAruc Hae Crurisrator of Oxford University Herbaria, Department of Plant Sciences, University of Oxford, South Parks Road, hecdloopnufsubelec f-usotpriav ercely vt.ie eLxwti neaenrs dn. buAemugtibnhe otrhrses o msnha oenuaulcsdhc r aippiamt gw eti toho fpt itrelexs pet anagtre e. given in the text should not be repeated in the tables SASA Ha Harrisrris consecutively. Line numbers on each page of text are given in the text should not be repeated in the tables Oxford,Druce X1Cu r3RB,ator UK.of O stephen.harris@plants.ox.ac.ukxford University Herbaria, Department of Plant Sciences, University of Oxford, South Parks Road, idhAeealll sp afugneld sf moinru fsrote rbvmeie nawuteimorsnb.e Arceludet achroolnyrs sea cnsuhdtoi vuceolydnc. Laiinsmeel nyt.ou mB pebrfeorsrse eon nt Acknowledgements should be brief with essential DDruceruc eCurator Curato ofr o Oxfof Oxrfodr Universityd Univers iHerbaria,ty Herbar Depaia, Dertmentpartm eofn tPlant of P lSciences,ant Scien Universityces, Unive ofrs iOxfoty of rOd,x Southford, S Parksouth Road,Parks OxfoRoadr,d, helpful for reviewers. Authors should aim to present COxford, Harwood X1 3RB, UK. [email protected] submittingiedaecahs p aanged your oifn tfeo xmanuscript,rtm araet ihoenlp cfulel fapleaseorrl yr eavni edreadw ecros .n thisAciuste hGuideloy.rs B she fotoourl de credits, withoutAcknow title(s)ledgem ofen tthes sh person(s).ould be brief with essential X1Oxford, 3RB, UK. X1 3RB,stephen.ha UK. [email protected]@plants.ox.ac.uk ideas and information clearly and concisely. Before Acknowledgements should be brief with essential CSIC RHaOr Landwood and Water, Private Bag 12, Hobart, Tasmania 7001, Australia. [email protected] Authorssubmittingaim to pcarefullrese yournt iyd.e manuscript,as and inform pleaseation c lreadearly thisand Guideconcise toly. credits, without title(s) of the person(s). CC Ha Harwoodrwood submitting your manuscript, please read this Guide to credits, without title(s) of the person(s). P CSIHoughtonRO Land and Water, Private Bag 12, Hobart, Tasmania 7001, Australia. [email protected] AuthorsBefore su carefullbmittingy .your manuscript, please read this Guide Proofs CSIROCSIRO Land Land and and W Wateater,r ,Private Private Bag Bag 12, 12, Hoba Hobart,rt, T Tasmaniaasmania 7001, 7001, Australia. Australia. chris.ha chris.harwood@[email protected] Authors carefully. TheP Houghton Lodge, Ballingham, Hereford HR2 6NN, UK. [email protected] to Authors careTfuypeslly. of contributions Proofs P Houghton Authors are respoPnroofssible for the accuracy of FEThe Putz Lodge, Ballingham, Hereford HR2 6NN, UK. [email protected] Types of contributions The Lodge, Ballingham, Hereford HR2 6NN, UK. [email protected] JTFS acceptsTypes rese ofarch contributions articles and review papers. submitteAd uwtohrokr, sin acrlued rinesgp cohnasnigbelse dfoonre t dhuer iancgc eudriatcinyg o. f Department of Botany, University of Florida, Gainesville, FL 32611, USA. [email protected]fl.edu Authors are responsible for the accuracy of ResearchJT aFrSti calccese aptsre rreespeoarchrts oafrtic orliesgi naandl r reesveiaerwc hpap anedrs . Fosur bpmapitetre dw iwtho rjok,i nint caluutdhionrgs hchipa,n ognees aduotnheo dr umriunsgt aecdcietipntg . FERM Putz Rowell JTFS accepts research articles and review papers. submitted work, including changes done during editing. contributeResearch toar ttheicle forests are rscienceeports andof o rtechnologiginal resye.a Reviewrch an d thFeo r epsappoenrs wibitihli tjyo ifnotr aaultlh coorrsrheispp, oonde eanucthe oarn md uhsat sa ctcheep t 4510FE Putz Gregg Road, Madison, WI 53705, USA. [email protected] Research articles are reports of original research and For paper with joint authorship, one author must accept FE Putz pacontributepers are fu toll-l theeng tforesth stat escience-of-the- aandrt r etechnologports thaty d. Reviewiscuss auththeo reitsyp toon ascibt iolint yb feohra lafl ol fc tohrere csponotnridbeunticneg a anudth hoars itnh e RMAJ Sinskey Rowell contribute to the forest science and technology. Review the responsibility for all correspondence and has the pupbalpisehres da rweo fruklsl.- lWenhgetnh ssutabtme-iottfi-tnhge r-aervti erwe ppoarptes rtsh, aatu dthisocruss s alla umattersthority pertainingto act on b etoh apublicationlf of the con oftr ithebut imanuscript.ng authors i n 451068-370A,RM GreggRowell Depa Road,rtment Madison, of Biolog WI y53705,, Massachusetts USA. [email protected] Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA. papers are full-length state-of-the-art reports that discuss authority to act on behalf of the contributing authors in RM Rowell arepu requestedblished wo rtoks consult. When stheubm Editoritting rinev iadvance.ew papers, authors Pralloo mattersfs will b pertaininge sent to t htoe publicationcorrespond iofng the au tmanuscript.hor to be [email protected] Sinskey Gregg Road, Madison, WI 53705, USA. [email protected] published works. When submitting review papers, authors all matters pertaining to publication of the manuscript. 4510 Gregg Road, Madison, WI 53705, USA. [email protected] are requested to consult the Editor in advance. chPerockoefsd w fiollr beer rsoenrst taon dth ere ctourrneesdpo pnrdoinmgp autlyt, huosru taol lyb e 68-370A,AJ Sinskey Department of Biology, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, are requested to consult the Editor in advance. Proofs will be sent to the corresponding author to be JRAJ V Sinskeyincent Preparation of research articles withinchec k48ed hours for e (notrror scounting and ret uweekendsrned pro andmpt commonly, usuall y USA.68-370A, [email protected] Department of Biology, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, checked for errors and returned promptly, usually Nicholas68-370A, School Department of the Enviof Biologronmenty, Massachusetts and Earth Sciences, Institute Duke of T echnologUniversity,, 77Box Massachusetts 90328, Durham, Avenue, NC 27708, Cambridge, USA. MA 02139, Preparation of research articles publicwithin holidays). 48 hours Only(not countingminimal alterationsweekends andare allowedcommon JRUSA. Vincent [email protected] Title pPargeparatione should ofinc resealuder chtit laer,ticles author(s) and within 48 hours (not counting weekends and common [email protected] [email protected] atpublic this stage. holidays). Only minimal alterations are allowed NicholasJR Vincent School of the Environment and Earth Sciences, Duke University, Box 90328, Durham, NC 27708, USA. Title page should include title, author(s) and public holidays). Only minimal alterations are allowed MJJR W Vingfieldincent Title page should include title, author(s) and at this stage. [email protected] School of the Environment and Earth Sciences, Duke University, Box 90328, Durham, NC 27708, USA. Abstract must be on a separate page. It should be at this stage. University of Pretoria, 0001 Pretoria, South Africa. Mike. Wingfi[email protected] Offprints Nicholas School of the Environment and Earth Sciences, Duke University, Box 90328, Durham, NC 27708, USA. a clear andAbs concisetract mu one-paragraphst be on a separ summaate pagery. I(lesst sho uthanld b e [email protected] Abstract must be on a separate page. It should be Offprints [email protected] a clear and concise one-paragraph summary (less than Offprints a clear and concise one-paragraph summary (less than SUBSCRIPTION RATES PER VOLUME results and major conclusion without the use of reference. coJTrrFeSs psuopnpdliiens ga caoumthploimr.e Tntharey PDF file to i tsh feo cro rpresrpsoonndainl g SUBSCRIPTION RATES PER VOLUME resultsK aenydw omrdajso (r 5c–o1n0c lwuosirodns wseitphaoruate tdh eb uy sceo omf mreafes)r eanrece . usceo rornelsyp aonndd icnagn naoutt hboer .c oTphied P, DdFis tfriilbeu itse dfo orr p peorstoenda l SUBSCRIPTION RIndividualATES PER VOLUMEInstitution ISSN: 0128 - 1283 results and major conclusion without the use of reference. caoutrhroers.p Tohned PiDngF au tchanonr.o tT bhee p oPsDtedF ofnil aen iys wfeobrs iptee wrsitohnoault SUBSCRIPTION RATES PER VOLUME intendedK efoyrw olirtdersa (t5u–r1e0 r wetorridevs asle. pTaor awteide bny cseoamrcmha, st)h aer e onu sea noyn wlye abnsdit ec awninthoot ubte pcoripoire dper, dimstirsisbiuotne df roorm p othste d Individual Institution DateISSN: of 0128issue: - October1283 2016 Keywords (5–10 words separated by commas) are upsreio ro pnelyr mianssdio cna fnronmot thbe Dcoirpecietodr,- Gdiesntreirbaul FteRdI Mor. Fpoors tehdis Individual Institution ISSN: 0128 - 1283 keinteywonrddes ds hforu lldit ebrea toutrhee rre thriaenva tlh. oTsoe waildreand sye aprrcehse, ntht e Doirne ctaonry-Gen webesriatle F wRIMith.o Fuotr pthisrio re perquemst,i splsieoanse f wroritme toh e Malaysia IndividualRM45.00 InstitutionUSD45.00 ISSN:Date 0128 of issue: - 1283 October 2016 intended for literature retrieval. To widen search, the orenq uaensty, pwleeabsesi wter itwe ittoh tohue te dpirtoior.r permission from the Date of issue: October 2016 ink theeyw otitle.rds should be other than those already present theDi editorector.-General FRIM. For this request, please write to Overseas Malaysia RM90.00RM45.00 USD90.00USD45.00 Date© Forest of issue: Research April Institute2019 Malaysia 2016 keywords should be other than those already present Director-General FRIM. For this request, please write to Malaysia RM45.00 USD45.00 in the title. the editor. MalaysiaOverseas RM100RM90.00 RM200USD90.00 © Forest Research Institute Malaysia 2016 in the title. the editor. OverseasOverseas USD100RM90.00 USD200USD90.00 ©© Fo Forestrest Resea Researchrch Institute Institute Malaysia Malaysia 2019 2016 Printed by Publications Branch, Forest Research Institute Malaysia, Kepong Printed by PublicationsPrinted Branch, by Reka For Cetakest Resea Sdn.rch Bhd., Institute Shah Malaysia, Alam Kepong Printed by Publications Branch, Forest Research Institute Malaysia, Kepong Journal of Tropical Forest Science 31(2): 162–174 (2019) Elaieb MT et al. https://doi.org/10.26525/jtfs2019.31.2.162174 COLLAPSE AND PHYSICAL PROPERTIES OF NATIVE AND PRE-STEAMED EUCALYPTUS CAMALDULENSIS AND EUCALYPTUS SALIGNA WOOD FROM TUNISIA

Elaieb MT1, Ayed SB1, Ouellani S2, Khouja ML1, Touhami I1 & Candelier K3, 4, *

1LGVRF, INRGREF, B.P. 10, 2080 Ariana, Tunisia 2Arrondissement des forêts, Nabeul, rue d’Alger 4029, Nabeul, Tunisia 3CIRAD, UPR BioWooEB, F-34398 Montpellier, France 4BioWooEB, Univ Montpellier, CIRAD, F-34398 Montpellier, France

*[email protected]

Submitted February 2018; accepted May 2018

Eucalyptus is the second major wood species used for Tunisian reforestation since 1957, and they are found around the country in several arboretums. Eucalyptus may be an interesting raw material to the Tunisian wood industry. The main obstacle to its industrial exploitation is its natural propensity to incur internal checking, collapse and a high transverse shrinkage during industrial drying process. This study focused on the physical and mechanical properties of reforested Eucalytus saligna and Eucalyptus camaldulensis from the north west of Tunisia. Moisture content, densities, shrinkages and mechanical properties were determined. Then, the impact of pre-steaming on the physical properties of modified wood was investigated. The results showed that both Eucalyptus possess low dimensional stability and mechanical properties compared to other Eucalyptus species from Tunisia, Morocco, Australia and Brazil. These wood characteristics were mainly due to their low density and sensitivity to collapse reactions, occurred during drying. Pre-steaming process reduced Eucalypts wood moisture content, changing the wood permeability and resulting in a residual collapse recovery and a decrease in wood shrinkage. Pre-steaming treated E. camaldulensis and E. saligna wood could be valuable as furniture and/or structural material without being submitted to moisture content variation.

Keywords: Collapse phenomenon, densities, mechanical properties, shrinkages, Tunisian Eucalyptus

INTRODUCTION

Eucalyptus, a native genus from Australia, belongs Eucalyptus wood remains mainly restricted to the Myrtaceae family and consist of 900 species to fuelwood or pulpwood because of limited and subspecies. It is one of the world’s most research on utilisation of wood; usage of the important and widely planted genera. It has Eucalyptus wood as material is very limited been introduced worldwide, including Tunisia, in Tunisia (Kabir et al. 1995). Eucalyptus is and mainly cultivated for its timber, pulp and considered as an invasive wood species that essential oils that possess medicinal properties needs to be economically valued. Despite the and therapeutic uses (Slimane et al. 2014, abundance of the species, only recently efforts Sebei et al. 2015). Eucalytus is also known for have been made towards its full utilisation. its melliferous characteristics in the production Marketing of Eucalyptus products has been of high quality honey. This fast-growing species hindered by lack of knowledge about its wood adapted very well to the Tunisian climate properties. However Eucalyptus has currently and was used to stabilise the costal dunes of become a subject of interest as raw material for northwest Tunisia, reduce erosion and protect wood composite panels in many tropical and the roadsides (Khouja et al. 2001, Dia and subtropical countries including Thailand, Chile, Duponnois 2010). Nowadays, Eucalyptus is the Brazil and Malaysia (Nacar et al. 2005). second major Tunisian reforestation wood species In the coming years, there will be a large covering 43,000 ha with an annual production of supply of Tunisian plantation timber for 3 m3 annum-1 ha-1, that gives a potential annual different segments of the forest market. Thus, wood production of 120,000 m3 (FAO 2012). the development of Eucalyptus solid wood However, the industrial use of these valuable products seems to be very promising at the

© Forest Research Institute Malaysia 162 Journal of Tropical Forest Science 31(2): 162–174 (2019) Elaieb MT et al. national level. The main challenge in developing During steaming, the growth stresses of wood high value-added products from Eucalyptus, can be released by means of microstructural such as remanufactured lumber, is the inherent reorganisation and thus, improve the quality difficulties in drying this species. Because of the of timber (Length and Kamke 2001; Severo et natural propensity to internal checking, collapse al. 2010; Kiemle et al. 2014). The wood is also and a high transverse shrinkage, lumber grade softened and growth stress is released as it reaches recovery after industrial drying is generally very the glass transition temperature during drying low. Unacceptable internal checking and collapse process. The wood ductility is increased due to after drying, reported in scientific literature for its improved viscoelastic properties, and finally several Eucalyptus wood species, confirmed that results in decreased drying cracks and collapse the commercialisation of Eucalyptus as solid wood (Calonego et al. 2010). products remains a great challenge (Northway The objectives of this study were (i) to 1995, Ananias et al. 2014). determine the physical and mechanical properties Therefore, a good understanding of the of reforested E. camaldulensis and E. saligna, (ii) relationship between collapse and wood to evaluate the impact of steam pre-treatment characteristics is needed to successfully process on the physical properties of modified wood this species into commercial solid wood and (iii) to identify the potential valorisations products. In addition, wood pre-treatment can of E. camaldulensis and E. saligna woods in be performed to improve wood properties and Tunisia. The results obtained could contribute limit checking and recover collapse. A variety towards the development of a database, useful of pre-drying treatments is discussed ranging for planting and processing. It will also be useful from storage in controlled environments, to for tree breeders and silviculturists to identify various wrapping types, pre-surfacing and pre- the properties that need improvement through heating. Cutting boards radially, rather than breeding selection or forest management tangentially, can also be regarded as a form of strategies. pre-drying treatment, as well as the application of an end-coating to freshly sawn logs. Pre- MATERIALS AND METHODS heating green timber prior to drying include steaming microwave treatment and pre-freezing, Wood selection and sampling and boiling (Ellwood 1953, Lee and Jung 1985, Haslett and Kininmonth 1986, Vermaas and In this study, 50 years old E. camaldulensisa and Bariska 1995, Zhang et al. 2011, Kong et al. 2017). E. saligna wood species were chosen to estimate the Such treatments result in increased permeability physical and mechanical properties of Eucalyptus and drying rate, which reduces wood drying time wood species from Tunisian reforestation. (Yang and Liu et al. 2018). A total of 3 trees of each wood species were The steaming of wood is a technique which collected from Zarniza arboretum, governorate has been employed for a variety of purposes in of Bizerte, regions of Sejnane (37°9' N, 9°7' E), the conversion and utilisation of forest products. from selected healthy trees, free from defects These include the facilitation of bending, and alteration, with almost perfectly straight alteration of colour, reduction in levels of growth trunk (Figure 1), according to Oger et Lecerq stress, improvement in dimensional stability, (1997). The Eucalyptus trees were selected as increase in permeability and improvement of representatives of their respective populations penetration by preservatives, facilitation of within the arboretum (tree dimeter of 40 cm at enzymatic hydrolysis, reconditioning of collapse- 1.30 m from the ground). prone species and the reduction in drying The climate of this northwest region of Tunisia time and rate during seasoning (Tieman 1929, is sub-humid with an annual rainfall of 927 mm Greenhill 1938, Perkitny et al. 1959, Campbell year-1. Average annual temperatures range 1961, Ellwood and Erickson 1962, MacKay from 14.9 °C to 18.5 °C. The average maximum 1971, Chen 1975, Chen and Workman 1980, temperature of the warmest month reaches more Shimizu et al. 1983, McGinnes and Rosen 1984, than 35 °C, and the average minimum of the Weik et al. 1984, Lee and Jung 1985, So 1985, coldest month is around 4 °C. The soil is poorly Barnes 1986, Cutter and Phelps 1986, Haslett developed in coastal dunes with leached brown and Kininmonth 1986, Donaldson et al. 1988,). forest in the mountains (Rejeb et al. 1996).

Tunisia Region of Bizerte

Zarniza Arboretum Governorat of Bizerte–Region of Sejnane [37°9' N, 9°7' E]

Figure 1 Location of the selected E. camaldulensis and E. saligna trees

Physical properties (mh – m0) EMC (%) = 100  (1) m 0 Wood samples used for the physical and mechanical characterisations do not differentiate Where mh is the green mass of the initial sample the sapwood and heartwood for selected trees, and m0 is the oven-dried mass of the wood because the sapwood/heartwood ratios are sample. different relating to the trees. The aim of this study is to valorise eucalyptus woods without Density and shrinkage separating sapwood and heartwood. Wood samples for physical and mechanical tests were To perform the physical tests, two wooden disks randomly chosen from the log, in order to of 50 mm thickness were cut at 1.40 m from observe the high results generally obtained in the ground for each selected tree; one disk to such characterisations (Figure 2). An overview determine the native wood properties and the of sampling performed on the different trees, other to evaluate pre-steaming process on wood for physical and mechanical tests, is shown on collapse phenomenon (Figure 2). Selection of Figure 2. the samples was similar for the two disks (in position, distance to pith and azimuth angle Relative humidity from the north) in order to obtain a good comparison between native and steamed wood To measure the humidity of trees, a wooden samples. To avoid errors during sampling, disk of 50 mm thickness was cut at 1.35 m from extreme cases such as excessively knotty trees, the ground for each selected tree (Figure 2). presence of reaction wood or slope grain were Then, wooden disks with dimensions of 20 × 20 × taken into account (ISO 4471 1982). From each 10 mm (along the grain) were cut from the disk, samples of 3 cm width, from bark to bark, samples, according to the repartition presented were cut in both directions; N-S and E-W (Figure in Figure 3a. Profiles of humidity repartition 3b). These samples were then cut into strips of within the wooden disks were drawn and moisture 2 cm thickness. In total, the density and shrinkage content calculated using the following equation: measurements (with and without steaming) were

Figure 2 Overview of Eucalyptus wood sample selection for the physical and mechanical analyses

(a)

North

(b)

North

Figure 3 Wood sample repartition for humidity (a), densities and shrinkages (b) determination tests

© Forest Research Institute Malaysia 165 Journal of Tropical Forest Science 31(2): 162–174 (2019) Elaieb MT et al. carried out on 165 samples for both wood species, Similar operations were used to determine E. camaldulensis and E. saligna. tangential (βt), radial (βr) and longitudinal (βl) shrinkages, using dimensional variation of the Density respective orientation.

Basic density, air-dried density (after conditioning Shrinkage with steaming in a climatic room at 20 °C and 65% RH) and oven-dried density (after conditioning in an In order to avoid the collapse phenomenon, each oven at 103 °C) (Dm12, Dm0) of the wood samples wood specimen was placed into a chamber with were determined according to International adjustable temperature and relative humidity Organization for Standardisation Standards using for progressive reconditioning until mass wood specimens of 20 × 20 × 10 mm (along the stabilisation of 16% moisture content. After grain) (ISO 13061-2 2014). The shape factor reconditioning, samples were autoclaved at (βt βr-1) was the ratio between tangential and 90 °C, 2.5 bars, for 30 min. Mass and 3-dimensional radial shrinkage. The densities were determined measurements were taken for each sample, prior by the gravimetric method (Haygreen and to drying at 103 °C until mass stabilisation. Bowyer 1996). Mass and 3-dimensional measurements of the

dried samples, and shrinkage (β0) [tangential m0 Db = (2) (βt0), radial (βr0), longitudinal (βl0) and Vh volumetric (βv0)] of the pre-streamed wood m samples were determined in the same way as for 0 Dm0 = (3) shrinkage without steaming (ISO 4469 1981). Vh -1 The shape factor (βt0 βr0 ) without collapse was also determined. m12 Dm12 = (4) V12 Comparison between shrinkage with or without steaming Where mh is the humid mass of the initial sample, m is the oven-dried mass of the wood 0 In order to evaluate the collapse phenomenon sample, and m and m are the oven-dried and 0 12 effect on wood shrinkage, Indicators of Collapse air-dried weight of the sample (g), respectively; -3 Recovery (IRC), in the three directions, were Db is the basic density of wood (g cm ), Dm0 is the -3 determined according to the following formula: oven-dried density of wood (g cm ) and Dm12 is the air-dried density of wood (g cm-3); V is the h (β – β ) green volume of the specimen (cm3), V is the v v0 0 IRC (βv) =  100 (6) βv oven-dried volume of the sample and V12 is the air-dried volume of wood sample. Similar operations were used to determine tangential IRC (βt), radial IRC (βr), IRC Shrinkage without steaming longitudinal (βl) shrinkages, and IRC(βt/βr), using dimensional variation of the respective Shrinkage (β) [tangential (βt), radial (βr), orientation. longitudinal (βl) and volumetric (βv) of the wood samples were determined according to Mechanical properties International Organization for Standardisation Standards using wood specimens of 20 × 20 × 10 mm (along the grain) (ISO 4469 1981). To perform mechanical resistance tests, three point bending (MOR) and compression tests The shape factor (βt βr-1) was the ratio between tangential and radial shrinkage. Volumetric were carried out for each of the selected wood shrinkage was measured using the following tree samples, and results were compared. A equation: universal mechanical test machine was used for the measurements. Samples were conditioned in a climate-controlled room with 65% RH and (Vh – V0) B =  100 at 22 °C, for the time required to stabilise the v V (5) h samples weights.

Bending test RESULTS AND DISCUSSION

Three point static bending tests were carried Moisture content out according to EN 408 (2010). The sample size was 300  20 mm  20 mm3 (L  R  T). Table 1 gives the minimal, maximal and The moving head speed and span length were average values of initial moisture contents of 0.09 mm s-1 and 260 mm, respectively. The load E. camaldulensis and E. saligna trees at 1.40 m deformation data obtained were analysed to from the ground, just after felling. On average, determine the modulus of rupture (MOR). The the initial moisture contents were 75.14% for tests were replicated on 20 samples from each E. camaldulensis wood and 87.27% for E. saligna selected Eucalyptus tree. wood, with respective minimal values of 59.80% and 36.95% and maximal values of 122.65% and Compression strength parallel to grain 116.71%. The moisture content of both eucalyptus wood species were relatively close. In comparison Compression tests were carried out according with other Eucalyptus species, moisture content to EN 408 (2010). Deviating from the norm, of E. camaldulensis and E. saligna were similar a reduced specimen size of 30  20  20 mm3 to those found in close geographical locations. (L x R x T) was used. The moving head speed Sahbeni (2014) found initial moisture content was 0.09 mm s-1 to ensure wood sample rupture values of 82.9% and 101.5% for E. bicostata and within 1.5 to 2 minutes. The load deformation E. coriacea woods from Souniat arboretum in data obtained were analysed to determine Tunisia. Even if these trees grew and were felled the modulus of rupture (MOR). A total of 20 in similar conditions with those of the present collapsed wood specimens per selected tree were study, moisture content comparison is still tested. needed. Elaieb et al. (2017) highlighted that moisture content of fresh felled E. loxophleba Statistical analyses and E. salmonophloia in Northeast Tunisia were 37.1% and 37.8%, respectively. However, the Statistical analyses (one-way analysis of variance) initial eucalyptus wood moisture can be largely using Fisher test and JMP 10.0.2 program variable. Ananias et al. (2014) found wood initial were performed. The effects of pre-streaming moisture content ranging from 132 to 200% for process on E. camaldulensis and E. saligna on different E. nitens trees from Las Mellizas site wood densities and shrinkages properties in Rucamanque farm, located near the city of were evaluated using ANOVA and Duncan’s Huepil in the eighth region of Chile. comparison test. Such analysis allows to class The studied E. camaldulensis and E. saligna results into several categories from A to C. woods was characterised by low initial moisture Systems which are not connected by the same content wood, when being felled. The results letter are largely different at the 5% level. showed that initial moisture content repartition

Table 1 Wet moisture contents average values: minimum and maximum values of E. camaldulensis and E. saligna woods

Moisture contents (%)

Wood species Min (%) Max (%) Average (%) SD (%) CV (%)

Eucalyptus 59.80 122.65 75.14 9.52 13.56 camaldulensis

Eucalyptus saligna 36.95 116.71 87.27 13.30 15.33

SD = standard deviation, CV = coefficient of variation

© Forest Research Institute Malaysia 167 Journal of Tropical Forest Science 31(2): 162–174 (2019) Elaieb MT et al. within the tree is different between E. camaldulensis Density and E. saligna woods. Figure 4 shows that the moisture content of E. saligna wood increased Air-dried density (Dm12) is commonly used to progressively from the pith to bark, whereas the compare different woods. Basic density (Db), opposite was observed for E. Camaldulensis wood. oven-dried density (Dm0) and air-dried density The mean values of moisture content for (Dm12) were measured on each E. camaldulensis E. saligna increased from 40% for the core and E. saligna wood sample. Average values, wood, 85% for transition wood area to 120% for maximal and minimal values of the different sapwood. On the contrary, the mean values of densities and coefficient of variations are moisture content for E. camaldulensis decreased presented in Table 2. -3 from 120% for the heartwood, 85% for transition Db, Dm12 and Dm0 were 0.639 ± 0.014 g cm , wood area to 60% for sapwood (Figure 5). 1.001 ± 0.014 g cm-3 and 0.772 ± 0.028 g cm-3 for According to literature, moisture content E. camaldulensis and 0.544 ± 0.070 g cm-3, 0.804 ± of fresh felled eucalyptus trees vary relating 0,019 g cm-3 and 0.739 ± 0,020 g cm-3 for E. Saligna, to the wood species. Elaieb et al. (2017) and respectively. Ananias et al. (2014) found similar trends in E. camaldulensis and E. saligna wood species is initial wood moisture content distribution within classified as heavy wood (Dm12 > 0.95) and from E. loxophleba, E. salmonophloia and E. nitens tree. mid-heavy wood (0.65 > Dm12 > 0.80) to heavy The results showed that initial moisture content wood (Dm12 > 0.95), respectively (Campredon of heartwood was considerably higher than 1967). transition and sapwoods. However, Fromm et Similar results were showed by a study al. (2001) highlighted that moisture content of conducted on 61 different Eucalytus tree species various Spruce and Oak trees decreases from early from Austalia (including E. camaldulensis, wood to late wood. E. platycorys, E. loxophleba and E. salmonophloia),

E. saligna E. camaldulensis

Figure 4 Moisture content repartition within E. camaldulensis and E.saligna trees

a: Juvenile wood b: External heartwood E. saligna c: Transition wood d: Sapwood

Figure 5 Location and definition of juvenile wood, heartwood, transition wood and sapwood

Table 2 Values of air dried, anhydrous and basic densities of E. camaldulensis and E. saligna

Wood species Min Max Average SD (%) CV (%)

-3 Dm12 (g cm ) Eucalyptus camaldulensis 0.993 1.017 1.001 0.014 0.014 Eucalyptus saligna 0.651 1.173 0.804 0,019 0,009 -3 Dm0 (g cm ) Eucalyptus camaldulensis 0.742 0.798 0.772 0.028 0.037 Eucalyptus saligna 0.413 0.673 0.738 0.020 0.127 -3 Db (g cm ) Eucalyptus camaldulensis 0.629 0.655 0.639 0.014 0.021 Eucalyptus saligna 0.363 0.687 0.544 0.070 0.130

SD = standard deviation, CV = coefficient of variation that Eucalytus wood basic density was between radial shrinkage (βr > 6.5%). According to the 0.690 and 0.940 g cm-3. In Tunisia, studies on literature, shrinkage of Tunisian E. camaldulensis wood characterisation were conducted in the and E. saligna woods are relatively close to those past two years on local Eucalptus wood species: of other Eucalyptus species such as E. loxophleba E. bicostata, E. cinerea, E. coriacea, E. maidenii, (βt βr-1 = 2.5) and E. salmonophloia (βt βr-1 = E. torquata and more recently on E. loxophleba 1.2) from Tunisia, E. globulus (βt βr-1 = 1.6) and E. salmonophloia woods (Selmi et al. 2014, from Morocco, E. citrodiora (βt βr-1 = 1.43) and Dridi 2015, Elaieb et al. 2017). The respective E. grandis (βt βr-1 = 1.64) from Brazil and basic densities of these wood species are 0.630, E. torquata (βt βr-1 = 1.27) from Tunisia (Segura 0.619, 0.711, 0.691, 0.870, 0.860 and 0.894 g cm-3. 2007, El Alami 2013, Sahbani 2014, Selmi 2014, Comparing the results from the current study Dridi 2015, Elaieb et al. 2017). with those from other Tunisians Eucalyptus species, E. camaldulensis and E. saligna seem to be Influence of steam pre-treatment on wood classified among the lightest TunisianEucalyptus properties woods. Finally, the large differences observed In some Eucalyptus species, wood shrinkage is between minimal and maximal values of Db, Dm0 often preceded by collapse phenomenon (Sesbou and Dm12 in E. saligna, could be explained by the and Nepveu 1978). Previous studies highlighted lower density of its juvenile wood than those of shrinkage variations of Eucalyptus wood species its sapwood. occurring above the wood fiber saturation point (Elaieb et al. 2017). This trend is probably due Shrinkage and shape factors to wood collapse reaction. Indeed, collapse of the cells during drying is commonly observed Shrinkage analyses shows the volumetric (βv), in certain timber species and is particularly tangential (βt), radial (βr), longitudinal (βv) pronounced in some members of the genus shrinkages and shape factor (βt βr-1) values of Eucalyptus (Berry and Roderick 2005). Collapse E. camaldulensis and E. saligna (Figure 6). These is defined as a form of irregular shrinkage results showed clearly that E. camaldulensis wood occurring above the fiber saturation point is more sensitive to moisture content variations (Assouad 2004, Tazrout et al. 2012). The most than E. saligna wood. Indeed, the average value of common explanation for collapse is capillary shape factor of E. camaldulensis (2.16) was higher tension. One of the causes of collapse is that the than those of E. saligna (1.81). The volumetric, cell walls cannot withstand the surface tension tangential and radial shrinkage average values of the water that is removed from the lumen of E. camaldulensis and E. saligna were 25.2 and of the fibers. On the other hand, macroscopic 22.1% (βv), 17.9 and 14.5% (βt), 8.1% and stresses arising in the wood during drying have 7.9% (βr), respectively. The E. camaldulensis been suspected to contribute to collapse, and and E. Saligna were classified as wood with a have been claimed by some workers to be the high volumetric shrinkage (βv > 13%), a high sole cause of the phenomenon (Greenhill 1938, tangential shrinkage (βt > 10%) and a high Stamm and Loughborough 1942). However,

E. Camaldulensis (with collapse) E. Camaldulensis ofter recontionning (without collapse) E. Saligna (with collapse) Indicator of collapse Shrinkages (%) E. Camaldulensis - indicator of collapse recovery recovery (ICR) E. Saligna - indicator of collapse recovery" 30 0,4

25 0,3 20

15 0,2

10 0,1 5

0 0,0 bv (%) bt (%) br (%) bt (%)/br (%) Shrinkages and rations for shrinkages

Figure 6 Average values of shrinkages and indicator of collapse recovery (ICR) in different orientation of E. camaldulensis and E. saligna before (with collapse) and after conditioning (without collapse) literature also acknowledges that collapse can strength of the wood (Campbell 1961, Rosen occur below the fiber saturation point (Almeida and Laurie 1983, Stubenvoll 1985) and increase et al. 2008). shrinkage, collapse and checking susceptibility Because of the natural propensity to internal during drying (Greenhill 1938, Kauman 1961, checking, collapse and a high transverse Liang 1981, Haslett and Kininmonth 1986). shrinkage, the Eucalyptus wood grade recovery, Nevertheless, there remains some disagreement after industrial drying, is generally low (Haslett with respect to the effect of pre-steaming on 1988). Anecdotal industrial experience suggests shrinkage and associated degrade. that Eucalyptus woods recovery after industrial Figure 6 shows that the mean values of drying could be as low as 20–30% for the thicker shrinkage in pre-streaming conditioned E. sawn sizes (Ananias et al. 2014). Unacceptable camaldulensis and E. saligna wood were lower internal checking and collapse after drying was than those of control specimens. However, the also reported in scientific literature for Eucalyptus dispersion of pre-treated woods shrinkages was wood (Ananias et al. 2009). This internal checking wider than those from control samples, which and collapse make the commercialisation of some does not highlight that steam pre-treatment has Eucalyptus solid wood products unviable for the a significant effect on wood physical properties. moment. In addition, according to IRCs indicators, E. The steaming of wood is a technique which saligna seems to be more sensible to collapse has been employed, for several years, for a variety than E. camaldulensis. In fact, whatever the of purposes in the conversion and utilisation wood direction, IRC values from E. saligna of forest products, including reconditioning of wood shrinkage was higher than those from E. collapse-prone species (Tieman 1929, Greenhill camaldulensis wood shrinkage. 1938). Through such steam treatment process, The results confirmed those obtained by wood absorb a small amount of moisture and previous studies on other Eucalyptus wood species. resume its initial shape which decreases the Ananias et al. (2014) highlighted that shrinkage effect of collapse. While such applications are before reconditioning and collapse of E. regnans valuable in improving wood utilisation, steaming wood showed highly significant increases in has disadvantages in that it may reduce the pre-steamed material with the exception of

© Forest Research Institute Malaysia 170 Journal of Tropical Forest Science 31(2): 162–174 (2019) Elaieb MT et al. sapwood, but shrinkage after reconditioning was cell lumens and relatively weak cell walls not significantly changed by pre-steaming for (Tiemann 1915). any data group. Campbell (1961) reported that steaming treatments of 2 to 4 h duration increase An increase in wood permeability together drying rates considerably without increasing the with a reduction in initial moisture content incidence of collapse. Further, in E. delegatensis should produce a reduction in collapse. This shrinkage before reconditioning and shrinkage decrease in moisture content in presteamed after reconditioning (2–4 h steaming) were not material was interpreted as reflecting changed significantly different from controls (Campbell permeability, likely responsible for the changed 1961). Unlike Haslett and Kininmonth (1986), level in residual collapse of pre-steamed material who observed unacceptable levels of internal (Ananias et al. 2014). or surface checking after 2 h of pre-steaming in Nothofagus fusca, Campbell (1961) observed Density less checking in E. obliqua after pre-steaming, although collapse was greater. In oak, Lee and As found by shrinkage properties determination, Jung (1985) found fewer end checks and slightly Db measurements of untreated and steam pre- less collapse and honeycombing after 4 h of pre- treated E. Saligna and E. camaldulensis woods steaming at 100 °C than in control specimens. showed higher sensitivity to collapse of E. saligna A number of hypotheses have been advanced to compared to those of E. camaldulensi. These account for changes in wood properties due to results were consistent with previous studies steam pre-treatment: which reported that collapse in Eucalyptus wood (i) Changes in chemical bonding among cell species tend to be higher when density of wood wall constituents or heat degradation are is lower (Kingston and Risdon 1961, Ananias et notably by lignin and hemicelluloses and al. 2009). Chafe (1986) recognised that collapse changes in wood extractives (Kauman 1961, increases proportionally to the increase in green Kininmonth 1971, Kubinsky 1971, Salud moisture content and reduction in basic density. 1976, McGinnes and Rosen 1984) Figure 7 shows that the mean loss in density (ii) In terms of liquid tension theory, the after steam pre-treatment of E. saligna (= 0.09) preconditions for collapse development is higher than for E. camaldulensis (= 0.01). are high impermeability, water-saturated Although statistically not significant, it was

Basic density (Db): with collapse

Basic density Basic density (Db): after steam treatment (without collapse) Indicator of collapse -3 recovery (ICR) in g cm Basic density: indicator of collapse recovery (ICR) 1 0.5 0.9 0.4 0.8 0.3

0.7 0.2

0.6 0.1

0.5 0

0.4 -0.1

0.3 -0.2

0.2 -0.3

0.1 -0.4

0 -0.5 E. camaldulensis E. saligna

Wood species

Figure 7 Average values of basic density (Db) and indicator of collapse recovery (ICR) of E. camaldulensis and E. saligna before (with collapse) and after conditioning (without collapse)

© Forest Research Institute Malaysia 171 Journal of Tropical Forest Science 31(2): 162–174 (2019) Elaieb MT et al. observed that steam pre-treatment slightly the species is classified as having high static decreased Db for both Eucalyptus wood species. bending strength and medium/high axial This decrease in density could be explained by compressive strength (Collardet and Besset the removal of extractives and low degradation 1998). of hemicelluloses. However, according to results, The lower mechanical properties of E. saligna disparity and lower mean value of IRC (positive and E. camaldulensis woods than those of other and negative values around 0.0), steam pre- Eucalyptus wood species from Tunisia could be treatment does not significantly affect the Db of due to their higher sensitivity to collapse and E. saligna and E. camaldulensis. lower density.

Mechanical properties CONCLUSION

The mechanical test results of E. saligna and E. The study focused on the physical and camaldulensis woods conditioned at a temperature mechanical properties of north-western E. of 20 °C and 65% RH are shown in Table 3. The saligna and E. camaldulensis woods issued from MOR average values of E. camaldulensis and E. Zarniza arboretum of the Sejnane region. The saligna woods were respectively 80.01 MPa and study was conducted on three trees each of 53.75 MPa in bending and 39.62 MPa and 31.72 Eucalyptus wood species from one geographical MPa in compression. Previous studies show that location. However, the results obtained cannot E. camaldulensis and E. saligna wood species can be generalised. Both species of Eucalyptus had be classified as having medium (55 MPa < σc < lower physical and mechanical properties 75 MPa) and low (σb < 75 MPa) static bending than other Eucalyptus wood in Tunisia and strength respectively, and low axial compressive other countries such as Brazil, Australia and strength (σc < 45 MPa) (Collardet and Besset Morocco, probably due to their sensitivity 1998). to collapse phenomenon. Due to collapse In comparison with literature, the results reactions, E. saligna and E. camaldulensis remain of E. saligna and E. camaldulensis woods had lowly exploited by lumber manufacturing lower bending and compression strength industry in Tunisia, for the development of properties than other Tunisian Eucalyptus species. commercial solid wood products. The steam Ghodhbéne (2014) characterised Tunisian E. pre-treatment technique was found effective cinerea and E. maidenii wood and found bending in recovering much of the collapse. It could strength values of 132.5 MPa and 107.2 MPa, be advantageously applied in Tunisia and and compression strength values of 54 MPa and other countries with Eucalyptus industries for 48.7 MPa. Sahbani (2014) found similar results better valorisation of ligneous products. This for E. bicostata with a compression strength value study showed that steam pre-treatment of of 50.40 MPa. Elaieb et al., (2017) showed that E. saligna and E. camaldulensis could improve MOR average values of Tunisian E. loxophleba wood properties and allow them to be used as and E. salmonophloia woods were 95.8 MPa and material to produce flooring, interior joinery, 97.2 MPa in bending test and 56.3MPa and 56.9 furniture, glulam and light frame for wood MPa in compression test, respectively. Thus, construction, in Tunisia.

Table 3 Mechanical properties of Tunisian E. saligna and E. camaldulensis woods: bending strength (σb) and compression strength (σc)

Wood species Min Average Max SDa CVb Eucalyptus camaldulensis 10.72 80.01 125.50 15.26 0.19 σb (MPa) Eucalyptus saligna 4.04 53.75 99.75 11.86 0.21 Eucalyptus camaldulensis 5.53 39.62 54.89 7.70 0.15 σc (MPa) Eucalyptus saligna 8.16 31.72 44.27 7.24 0.08

astandard deviation, bcoefficient of variation

REFERENCES de Thuya. PhD thesis. Sciences du Bois, Université Toulouse Mohamed V– Agdal, Maroc. Almeida G, Assor C & Perré P. 2008.The dynamic of Elaieb MT, Ben Rhouma S, Khouaja A, Khorchani A, Touhami shrinkage-moisture content behavior determined I, Khouja ML & Candelier K. 2017. Some Physical during drying of microsamples for different kinds and mechanical characterisation of Tunisian planted of wood. Drying Technology 26: 1118–1124. Eucalytus loxophleba and Eucalyptus salmonophloia Ananias RA, Diiaz C & Leandro L. 2009. Preliminary study woods. Journal of Chemical, Biological and Physical about the shrinkage and collapse in Eucalyptus Sciences Section B 7: 537–549. nitens [In Spanish]. Maderas: Ciencia y Tecnologia 11: Ellwood EL. 1953. Properties of beech in tension 251–262. perpendicular to the grain and their relation to Ananías RA, Sepúlveda-Villarroel V, Pérez-Peña N et al. 2014. drying. Forest Product Journal 3: 202–209. Collapse of Eucalyptus nitens wood after drying Ellwood EL & Erickson RW. 1962. Effect of presteaming on depending on the radial location within the stem. seasoning stain and drying rate in redwood. Forest Drying Technology 32: 1699–1705. Products Journal 12: 328–332. Assouad A. 2004. Modélisation du séchage sous vide EN 408. 2010. Timber Structures — Structural Timber and Glued discontinu du bois de chéne: intégration d’un critère Laminated Timber—Determination of Some Physical de qualité mécanique. Phd thesis, Science du Bois, and Mechanical Properties. ICS 91.080.20. European Université Bordeaux 1, Bordeaux. Committee for Standardization, Brussels. Barnes HM. 1986. Treatment of peeler cores with water Fao. 2012. Les Ressources Génétiques Forestières en Tunisie. Projet dispersible preservative formulations. Wood and Fiber tcp/rab/3303 — baby 2, rapport annuel. Etat des Science 18: 397–412. Ressources Génétiques Forestières en Afrique du Berry LB & Roderick ML. 2005. Plant-water relations and Nord. Roma, FAO Editions. the fiber saturation point.New Phytologist 168: 25–37. Fromm JH, Sautter I, Matthies D, Kremer J, Schumacher P Calonego FW, Severo ETD, Cunha AR & Gaia DC. 2010. Use & Ganter C. 2001. Xylem water content and wood of glass transition temperature for stabilization of density in spruce and oak trees detected by high- boards cracks of Eucalyptus grandis. Anais da Academia resolution computed tomography. Journal of Plant Brasileira de Ciências 82: 1–7. Physiology 127: 416–425. Campbell GS. 1961. The value of presteaming for drying Ghodhbene O. 2014. Etude des propriétés mécaniques de some collapse susceptible Eucalypts. Forest Products base du bois d’Eucalyptus cinerea d’Eucalyptus maidenii Journal 9: 343–347. dans l’arboretum de Souniat, notamment l’étude Campredon J. 1967. Cours de technologie du bois. Feuilles du collapse. MSc thesis. Licence Appliquée en documentaires. The Engineers of The Rural Protection de l’Environnement, Université de Tunis Engineering, Water and Forests (IGREF), Feuilles El Manar, Rommana. documentaries, Tunis. Greenhill WL. 1938. Collapse and its removal: Some Chafe SC. 1986. Radial variation of collapse, volumetric recent investigations with Eucalyptus regnans. CSIRO shrinkage, moisture content and density in Eucalyptus Australian Society of Forest Production 24: 1–32. regnans F. Muell. Wood Science and Technology 20: Haslett AN & Kininmonth JA. 1986. Pretreatments to hasten 253–262. the drying of Nothojagus Jusca. New Zealand Journal of Chen PYS. 1975. The effect of steaming time and temperature Forestry Science 16: 237–246. on the longitudinal permeability of black walnut. Haslett A. 1988. Properties and Utilization of Exotic Speciality Wood and Fiber Science 7: 222–225. Timbers Grown in New Zealand. Part V: Ash Eucalypts Chen PYS & Workman ECJR. 1980. Effect of steaming on and Eucalyptus nitens, FRI Bulletin No. 119. Scion, some physical and chemical properties of black Rotorua. walnut heartwood. Wood and Fiber Science 11: 218–227. Haygreen JG & Bowyer JL. 1996. Forest Products and Wood rd Collardet J & Besset J. 1988. Les bois commerciaux et leurs Science. 3 edition. Iowa State University Press, Iowa. utilisations. Bois résineux (conifères). Direction du ISO 4471. 1982. Wood: Sampling Sample Trees and Logs for Centre Technique du Bois et de l’Ameublement Tome 1: 277. Determination of Physical and Mechanical Properties of Cutter BE & Phelps JE. 1986. High-pressure steam drying: Wood in Homogeneous Stands. ICS 19.040. International effects on permeability. Forest Products Journal 36: Organization for Standardization, Geneva. 19–20. ISO 13061-2. 2014. Physical and Mechanical Properties of Wood Dia A & Duponnois R. 2010. Le Projet Majeur Africain de — Test Methods for Small Clear Wood Specimens. Part la Grande Muraille Verte: Concepts et Mise en Œuvre. 2: Determination of Density for Physical and Mechanical Mayenne; IRD Editions. Tests. ICS: 79.040, International Organization for Donaldson LA, Wong KKY & Mackie KL. 1988. Ultrastructure Standardization, Geneva. of steam exploded wood. Wood Science and Technology ISO 4469. 1981. Wood - Determination of Radial and Tangential 22: 103–114. Shrinkage. ICS 19.040. International Organization for Dridi H. 2015. Caractérisation physico-mécaniques du bois Standardization, Geneva. d’Eucalyptus torquata dans l’arboretum de Hinchir Kabir MF, Bhattacharjee DK & Sattar MA. 1995. Strength Naam. MSc thesis. Licence Appliquée en Protection properties of and drying characteristics of Eucalyptus de l’Environnement, Université de Tunis El Manar, wood grown in Bangladesh. Thai Journal of Forestry Tunisie. 14: 103–109. El Alami S. 2013. Contribution à la caractérisation physique Kauman WG. 1961. Effect of thermal degradation on et mécanique et la valorisation par le sechage du bois shrinkage and collapse of wood from 3 Australian de la forêt marocaine: Cas des eucalyptus et la loupe species. Forest Products Journal 11: 445–452.

Khouja ML, Khaldi A & Rejeb MN. 2001. Resultants of the Salud EC. 1976. Comparison of the quality and composition Eucalyptus Introduction trials in Tunisia. Ed. Centro of the wood extractives of unsteamed and steamed Propozione Pubblicita, Taormina. apitong (Dipterocarpus grandiflorus Blanco). Pterocarpus Kiemle SN, Zhang X, Esker AR, Toriz G, Gatenholm P & 2: 123–134. Cosgrove DJ. 2014. Role of (1,3)(1,4)-β-glucan in Sebei K, Sakouhi F, Herchi W, Khouja ML & Boukhchina cell walls: Interaction with cellulose. Biomacromolecules S. 2015. Chemical composition and antibacterial 15: 1727–1736. activities of seven Eucalyptus species essential oils Kingston RST & Risdon CJE. 1961. Shrinkage and density leaves. Biological Research 48: 1–7. of Australian and other South-West Pacific woods. Segura TES. 2007. Étude des propriétés viscoélastiques du CSIRO Australian Society of Forest Production 13: 6–7. bois saturé de peuplier et d’Eucalyptus. MSC thesis, Kininmonth JA. 1971. Effect of steaming on the fine structure Université de São Paulo (USP), Brésil. of Nothofagusfusca. New Zealand Journal of Forestry Selmi A. 2014. Etude des propriétés physiques de base Science 1: 129–139. du bois d’Eucalyptus cinerea d’Eucalyptus maidenii Kong LL, Zhao ZJ, He ZB & Yi SL. 2017. Development of dans l’arboretum de Sounia. MSc thesis. Licence schedule to steaming prior to drying and its effects Appliquée en Protection de l’Environnement, on E. grandis × E. urophylla wood. European Journal of Université de Tunis El Manar, Rommana. Wood and Wood Products 761: 591–600. Sesbou A & Nepveu G. 1978. Variabilité infra spécifique du Kubinsky E. 1971. Influence of steaming on the properties retrait avec collapse et de la densité du bois chez of Quercus rubra L. Wood. Holzforschung 25: 78–83. Eucalyptus camaldulensis. Annals of Forest Science 35: Lee HW & Jung HS. 1985. Effect of pre-streaming on the 237–263. drying rate and drying effets of Quercus acutissma. Severo ET, Calonego FW & Matos CA. 2010. Lumber quality Mogjaegonghak 13: 3–13. of Eucalyptus grandisas a function of diametrical Length CA & Kamke FA. 2001. Moisture dependent softening position and log steaming. Bioresource Technology behavior of wood. Wood and Fiber Science 33: 492–507. 101: 25–45. Liang SZ. 1981. The variability of swelling and shrinkage Shimizu K, Sudo K, Nagasawa S & Ishihara M. 1983. Enzymatic of vermillion wood and the technique of high- hydrolysis of wood. VII. Enzymatic suspectibility temperature drying. Journal of Nanjing Technological of autohydrolysed woods. Journal of the Japan Wood College of Forest Products 1: 1–15. Research Society 29: 428–437. Mcginnes EAJR & Rosen HN. 1984. Macroscopic and Slimane BB, Ezzine O, Dhahri S & Ben Jamaa ML. 2014. microscopic analyses of color changes of wood Essential oils from two Eucalyptus from Tunisia pressure steam-dried above atmospheric pressure. and their insecticidal action on Orgyia trigotephras Wood and Fiber Science 16: 48–56. (Lepidotera, Lymantriidae). Biological Research 47:1–29. Mackay JFG. 1971. Influence of steaming on water vapour So WT. 1985. Effects of boiling, steaming and chemical diffusion in hardwoods. Journal of Wood Science 3: treatment on solid wood bending of Quercus acutissima 156–160. and Pinus densiflora. Mogjaegonghak 13: 19–62. Nacar M, Hiziroglu S & Kalaycioglu H. 2005. Some of Stamm AJ & Loughborough WK. 1942. Variation in shrinking the properties of particleboard panels made from and swelling of wood. Transactions of the American eucalyptus. American Journal of Applied Sciences Special Society of Mechanical Engineers 64: 379–386. Issue: 5–8. Stubenvoll A. 1985. Effect of steaming on the bending Northway RL. 1995. Techniques to monitor drying stresses modulus of elasticity and bendingstrength of beech and dimensional changes in timber from plantation- and robinia. Faipar 35: 264–271. grown eucalypts for kiln schedule development and Tazrout M, Abadlia MT & Mothe F. 2012. Étude de la kiln control. Maderas: Ciencia y Tecnologia 4: 148–154. récupération du collapse par microdensitométrie Oger R & Leclercq A. 1977. L’échantillonnage appliqué avec du bois d’Eucalyptus camaldulensis Dehn à la détermination des propriétés physiques provenant de Beghla en Algérie. Bois et Forêts des et mécaniques du bois. Bulletin des Recherches Tropiques 312: 63–75. Agronomiques de Gembloux 12 : 331–338. Tiemann HD. 1915. Principles of kiln drying. Lumber World Perkitny T, Lawniczak M & Marciniak H. 1959. Uber den Review 28: 23–28. Einflulss des Dampfens auf den Quellungsdruck des Tiemann HD. 1929. How to restore collapsed timber. The Holzes. Holz Roh-Werkstoff 17: 54–61. Lumber Worker 5: 37–44. Rejeb MN, Khaldi A, Khouja ML, Garchi S, Ben Mansoura A Vermaas HF & Bariska M. 1995. Collapse during low & Nouri M. 1996. Guide pour le choix des espèces temperature drying of Eucalyptus grandis W. Hill de reboisement: Espèces forestières et pastorales. and Pinus sylvestris L. Holzforchung und Holzverwertung INGREF: 137. 47: 35–40. Rosen HN & Laurie SE. 1983. Mechanical properties of Weik BB, Wengert EM, Schroeder J & Brisbin R. 1984. conventionally kiln-dried and pressure steam dried Practical drying techniques for yellow-poplar S-D-R yellow-poplar and red oak. Forest Products Journal 33: flitches.Forest Products Journal 34: 39–44. 50–52. Yang L & Liu H. 2018. A Review of Eucalyptus Wood Collapse Sahbani M. 2014. Etude de quelques caractéristiques and its Control during Drying. BioResources 13: physiques et mécaniques du bois d’Eucalyptus. 2171–2181. Cas d’Eucalyptus bicostata et d’Eucalyptus coriacea Zhang YL, Miao P, Zhuang SZ, Wang XM, Xia JW & Wu LM. dans l’arboretum de Souniat. MSc thesis. Licence 2011. Improving the dryability of Eucalyptus by pre- Appliquée en Sciences et Techniques Forestières, microwave or pre-freezing treatment (in Chinese). Institut Sylvo-Pastoral de Tabarka, Tabarka. Journal of Nanjing Forestry University 35: 61–64.