Dobhal, ShivaniSingh, K. GullamaniTayeng, Tisu2022-10-152022-10-152022-09-09https://krishikosh.egranth.ac.in/handle/1/5810188938Thepresentstudy entitled,“Improvementthroughselectionforgrowthandwoodquality inGmelina arboreaRoxb.” was carried out during 2021-22 in twenty five years old plantation of G.arboreaRoxbatSimenChapori(latitude 27044’18’’N-27044’25”N,longitude94051’40”E-94052’03’’E, elevation140m-244m amsl (Arunachal Pradesh) and in Forest Biology and TreeImprovementlaboratory of College of Horticulture and Forestry, CAU, Pasighat (ArunachalPradesh). The objective of the study was to screen and select plus tree through index method ofselection for improvement in growth parameters and to evaluate selected plus trees for inter and intratrees variability based on wood anatomical, woodspecificgravityandfibrederivative propertiesA total of 61 candidate plus trees (CPTs) were enumerated for their morphometrictraits viz, Height, CBH, DBH, CD, Straightness, Branching behaviour and bark thickness. Out of61 CPTs, 28 Plus trees (PTs) were screened based on their total index score value>average.Forwoodanatomicalvariabilityanalysisincrementwoodcoresalongtreeradiusat1.37mheight from each selected trees were sampled. Each increment cores were further divided intothree equal sections (from pith to bark). In this manner, a total of 252 wood increment coresweresubjectedforanatomicalandfiberderivativestudies.Woodanatomicalparametersincluded 5 anatomical and 6 fibre properties viz. fibre length (FL), fibre diameter (FD), lumendiameter (LD), wall thickness (WT) and specific gravity (SG); and fibre derivative properties,Runkel ratio (RR), flexibility coefficient (FC), slenderness ratio (SR), Luce’s shape factor (LSF),wall coverageratio(WCR) andrigiditycoefficient(RC)wasestimated. On the basis of indexmethod of selection,highestimprovement percentwas observedinCBH(12.38%)followedbyheight(11.22%).MaximumselectiondifferentialwasobservedinCD(5.31)followedbyheight(2.75).ThepopulationofselectedplustreeswerefurtherclusteredbyToucher’smethodfordiversityanalysis.ClusterVandClusterXhadthemaximuminterclusterdistance(d2=1096632.0)and minimuminterclusterdistancewasobservedbetweenClusterIandClusterV(d2=40617.9).ClusteranalysisbasedongrowthtraitsrevealedthatClusterV,ClusterVI,ClusterVIIandClusterVIIIwithgenotypesCHF/AP/GA/29,CHF/AP/GA/48, CHF/AP/GA/60 and CHF/AP/GA/38 in each clusters were divergent respectively.Fromthecorrelationstudies,itwasobservedthatheightandCBHandDBHandCDhassignificantandpositivelycorrelatedingrowthtraits. Analysis of variance based on wood fibre properties revealed that core sections (pith tobark) showed substantial variations across core section. The pooled mean values recordedforFL(1369.95μm),FD(39.43μm),LD(30.28μm),WT(4.57μm),RR(0.31),SR(35.23),FC(76.62), RC (0.12), LSF (0.26) and WCR (0.23μm). It was observed that among 11 fibres traitsincluding fibre derivative properties selected genotypes values increases radially from pith tobark except in FC. The result of the study showed that out of twenty-eight genotypes, fourgenotypes were found to be the divergent for growth traits, however, genotypes CHF/AP/GA/9and CHF/AP/GA/49werefound to be the most productivein wood traits including pulp andpaperproperties.EnglishThesis