Thesis (MSc)--Stellenbosch University, 2001. / ENGLISH ABSTRACT: A total of 125 trees covering a range of tree diameters were selected from a progeny trial of open pollinated progenies of Pinus patuia for studies on trait assessments and evaluation for branching characteristics. The trees comprised 25 families with each family having five trees. The trial was 36 months old and was located in a relatively high yielding site in Block B at Usutu in Swaziland. The height, diameter, the number of branches and the number of whorls of each tree were recorded. In addition to these objective assessments on individual trees, trees were given scores for branch size, branch angle, stem straightness and crown coarseness. For the detailed branch evaluation study, trees were marked at 20% and 50% of the total tree height and the section between this range was used. On this section the internode length, the internode diameter, the branch diameter and the branch angle were measured. It was observed that the objective assessments correlated strongly with the scoring (subjective) and it was further recommended that a closer look is needed to consider the prospects of employing subjective assessments as a means of reducing costs incurred during objective trait assessments. The branch sampling study revealed that there were prospects of assessing fewer branches on a tree as opposed to assessing all the branches, which could be a major contributor to the high costs in tree breeding. All three methods used to explore this prospect showed some reasonable agreement with each other in this regard. The best sample is to measure all branches in whorl one and two above 20% tree height followed by sampling four branches, two from the first whorl and two from the second whorl. There is a slight difference in these two samples as determined by the coefficents of determination. The study recommends that four branches can be used in the assessment of the branch angle and the branch diameter on the trees when one considered the cost elements. / AFRIKAANSE OPSOMMING: In 'n ondersoek na die evaluering van tak-eienskappe van Pinus patuIa is 125 bome in 'n oop-bestuifde nageslags-proef geselekteer. Hierdie bome is gekies met groottes wat strek oor 'n wye verskeidenheid van borshoogtedeursnee. Die ondersoek het 25 families ingesluit met elke familie wat uit 5 bome bestaan. Die proef is 36 maande oud en is gevestig in'n hoë produksiearea van Blok B, Usutu, in Swaziland. Die hoogte, borshoogtedeurnee, aantal takke en die hoeveelheid takkranse is gemeet en bepaal by elk van die 125 bome. Verder is punte toegeken vir takgrootte, takhoek, graad van stamkromming en kroon grofheid. In die gedetailleerde tak-evaluasie is slegs die stamseksie tussen 200!o en 50% van die totale boomhoogte gebruik. Op hierdie seksie van die boomstam is die internodale afstand, die internodale deursnee, die takdeursnee en die tak-hoek gemeet. Uit die gegewens wat versamel is, het dit geblyk dat daar sterk korrelasie tussen die objektiewe evaluasie en die subjektiewe evaluasie (met gebruik van kwalitatiewe eienskappe op 'n tellingbasis) verkry is. Hieruit word aanbeveel dat meer aandag aan subjektiewe evaluasie (as vereenvoudigde evaluasieprosedure) bestee word om kostes te besnoei. Die gedetailleerde tak-evaluasiestudie het aangedui dat die gebruik van slegs 'n paar takke, in teenstelling met al die takke op die boom, voordelig kan wees om verdere kostes van evaluasie te bespaar. Die beste monster is om al die takke in die eerste twee takkranse bo 20% boomhoogte te meet. Daarnaas is die beste monster die eerste twee grootste takke in die eerste en tweede takkranse. Die gebruik van hierdie kleiner monsters vir evaluering het in sekere opsigte nog beter voorspellings gelewer as in die gevalle waar al die takke in die eerste takkrans gebruik is. Die studie b aan dat In betroubare monster van vier takke (twee uit elke takkrans bo 20% boomhoogte) geneem kan word om die boom se takeienskappe te be-oordeel.