Wood density and fibre morphology of Pinus caribaea var. hondurensis Barrett and Golfari from Fiji were studied in relation to pulp properties. Densitometry and anatomical measurement were examined using billets taken from a sub-sample of 20 12-13 year old trees, chosen as representative of the density range observed in a first sample of 120 trees. Pulping properties were examined at the Tropical Products Institute, London. Considerable differences between trees were found in all characteristics examined and all characteristics showed fairly marked tree to tree variation in development patterns in a radial direction though differences along the stem were small. Between ring differences for ring width, density, cell wall thickness and cell wall percentage were considerable but cell lumen width and cell diameter were fairly constant between rings. Tree morphological characteristics cannot be used for estimation of whole tree values for any of the features studied. Whole tree values for wood characteristics can be accurately predicted from samples taken at 5% height level. Characteristic patterns of wood density distribution classes and tracheid cross sectional dimension classes were found, resembling chi-squared, binomial and bimodal curves. Wood properties could be used to predict pulp yield percentage and tear factor accurately by simple regression, less precise estimates could be obtained in this way, for burst factor and breaking length, which were best indicated by cell wall thickness. The best indicators of variation in pulp total yield percentage were chemical properties. Density was the best indicator of tear factor whilst cell wall thickness was the most important factor in explaining burst factor and breaking length. Growth rate showed negative correlation with wood density; trees of lower density produce paper of a strength generally at least equal to that obtained with higher density trees, though tear factor was lower. Thus the selection of trees for fast growth rate is suggested as the most profitable course. The most desirable selection criteria are listed.

A mathematical model is presented for growth and yield of Cupressus lusitanica, Pinus radiata and Pinus patula growing in the highlands of Kenya, Uganda, Tanzania, and Malawi, based on permanent sample plot data. Stand dominant height is predicted by a log/reciprocal function of age and site Index. Initial diameter distributions are predicted by spacing at a dominant height of 7 metres and generated by a Weibull function. The diameter distribution of stems is simulated by a vector of 10 cumulative percentiles which are incremented over each growth period by a function based on relative basal area, dominant height, and tree dominance. Relative basal area is defined in terms of maximum basal area, which. itself is modelled by a Chapman-Richards function in terms of dominant height. Dominance is defined as the ratio of tree diameter to stand dominant diameter. Thinnings are simulated by a novel and efficient algorithm that takes some trees from all size classes but is biassed to the smaller stems, although the degree of bias varies with thinning intensity. Tree volumes are determined by a new tree taper function using a quadratic/cosine model of relative height and relative sectional area. An algorithm is presented based on the taper function for subdividing stand volume into classes determined by log diameter and length. Validation of the model showed that diameter increment and height prediction, as well as the thinning routine were very accurate over a wide range of stand conditions; but the taper function was biassed towards underestimation of larger sizes and required re-fitting by a suitable method of weighting, whilst the initial distribution estimation required an additional variable to account for some of the variation occurring in practice. Two computer programs and a subroutine library are presented in the Appendix. VYTL-2 is a program to produce variable yield tables; whilst PYMOD is a forecasting program for multicompartment forests.

Using data from temporary and permanent sample plots various functions describing different stand parameter relationships were developed for P. patula growing at Sao Hill, Southern Tanzania. The functions include equations for: a Weibull diameter distribution, a single tree diameter increment, stand basal area, mortality, and height/diameter relationship. Also site index curves based on a Chapman-Richards equation were constructed, and a compatible taper/volume estimation system was developed. The functions were integrated into a micro-computer model SIMUL in BASIC language. The model is capable of estimating saw-log and pulp-log volumes for different spacings and thinning regimes. Inputs to the model are: initial and simulation ages, stocking/ha, stand mean and standard deviation of diameter at breast height (dbh), basal area/ha, site, dominant height, and minimum diameters and lengths of saw-logs and pulplogs. The output gives yearly values/ha for; stocking, total volume, total basal area/ha, means for dbh, height and tree volume, mortality number and volume/ha, and volumes of saw-logs and pulplogs as out-turn from subsequent thinnings and the final clearfelling. Validation with independent data showed that the model works satisfactorily. Experimental runs at different levels of spacing and thinning regimes suggested that the current thinning schedule at Sao Hill is too heavy and results in loss in total volume. The highest utilizable volumes (mainly saw logs) may be obtained by using the current spacing of 2.7 x 2.7m followed by one thinning at the age of about 11 years to leave about 1100 stems/ha, with a rotation age of 25 years. This treatment is also justified economically giving the highest present value at an interest rate of 3%. The current rotation age 15 years (unthinned) for the production of pulp wood only is ideal to produce the required dimensions (10-20 cm diameter) of pulp logs. SIMUL is recommended for use at Sao Hill and the technique may be adopted for other plantation species in Tanzania with new parameter estimates for the equations.

The nature and occurrence of mycorrhizal associations, with particular reference to the anatomy, carbohydrate physiology, plant mineral nutrition and occurrence of ectomycorrhizae (ECM), is discussed. The ecology and forest relations of <i>Picea sitchensis</i> - the Sitka spruce concludes the literature review. Identification of areas of good and poor Sitka growth on related soils and the quantification of their ECM status, investigation of the effect of phosphate addition to Sitka seedlings in pots, subsequent and changes to their ECM status, and the effects of soluble aluminium on phosphate nutrition of Sitka seedlings, the collation of results and relation back to the field situation were carried out as experimental work. Field sites were identified and described in terms of geology, soils, field ECM status, forest productivity and nutrient status. Three pot experiments were carried out. The 1st investigated the effects of phosphate application on ECM Sitka seedlings in soil from the field sites; the 2nd investigated the effects of phosphate application to ECM and non-mycorrhizal (NM) seedlings in compost; and the 3rd investigated the affects of application of Al-citrate to ECM and NM seedlings in compost containing high and low levels of phosphate. The results obtained were described and discussed in the context of a model of the factors affecting plant response to the soil environment. The field ECM development representing a considerable drain on the carbohydrate economy of the field sites was to some extent duplicated in the greenhouse. The possible decrease in importance of this drain was illustrated by phosphate application, but was increased by addition of Al-citrate. A previously unrecorded ECM-enhanced uptake of Manganese was reported. The importance of phosphate in the soils under discussion was emphasised, and possible further work suggested.

The study examined the damage caused to naturally regenerated Scots Hue by red deer and certain other mammals and birds, and its effect on the development of natural woodland. Investigations were concentrated in three areas, which carried high, moderate and low stockings of red deer, in the upper Dee valley in Aberdeenshire. Most marked trees, above 2-3 inches in height, suffered some degree of browsing during the study period. Simulated browsing experiments, supported by data from study plots, showed that survival and growth of trees decreased with increased browsing severity. Severe browsing had less serious effects on large trees than on small ones. In response to browsing, new shoots were produced from large overwintering buds in larch and birch but, in pine, shoots were normally produced from dormant buds, which required a growing season in which to develop before vigorous growth was possible. Where browsing pressure was high, the deciduous species generally survived longer than pine because they were not subjected to severe winter damage and they probably suffered less interruption to their photosynthetic activity as a result of their more rapid response to browsing. Tree browsing records and observations in study plots confirmed that numerous mammals and birds browsed pine. Most species, however, had relatively little effect on woodland development. Densities of rabbits and mice were very low, and pine material was probably only a minor constituent in the diet of mountain hares and red grouse. Browsing by jred and roe deer could not be distinguished consistently, but the satisfactory development of natural regeneration in exclosures, where red deer were largely excluded but roe were common, suggests that roe densities of up to approximately 1*24 acres are too low to exert a serious effect on woodland development. Sheep commonly browsed pine in winter, but most animals were nonnally taken off the hill from November to May. Browsing on young regeneration by capercaillie was apparently restricted to pine of .4-30 inches high, which were few where capercaillie were numerous. Disbudding by black grouse was widespread but, in these areas, their browsing never affected tree survival and rarely prevented the addition of some increment during the year of damage. Browsing by red deer had the most serious effects on the survival and development of pine. Browsing had relatively little effect on tree development where red deer densities were approximately 1*150 acres in winter (Cambus O'May); but it caused mortalities in trees below 30 inches high and reduced growth in trees below 60 inches, in areas where winter densities approximated to 1:60 acres (Bush Lawsie). Densities of It10 acres (Meall Alvie) totally prevented the development of woodland. A comparison of tree browsing incidence with indices of red deer abundance in study plots, and a detailed examination of the conditions under which these trees were browsed, showed that whilst intensity of pine browsing is related to deer density, the two are not directly proportional to each other. The relationship is modified by environmental factors of climate and the habitat, notably the choice and availability of tree material and of the alternative vegetation which comprises the deer*a diet. Observations on feeding, rumen content analyses and comparisons of browsing intensity on trees of different origin and appearance showed that red deer browsed certain species of vegetation, and trees of certain sizes and condition, selectively. Habitats which contained large quantities of preferred vegetation and low quantities of acceptable tree material, therefore, suffered less pine browsing proportional to their use by deer than habitats in which these conditions were reversed. Deer abundance was invariably a significant factor in the amount of tree browsing which a habitat sustained, however, because increased browsing intensity removed more preferred vegetation and acceptable tree material and forced the deer to browse more heavily on less acceptable trees. Tree mortality from trampling had little effect on woodland development because damage was confined to very small trees in areas where regeneration was profuse. Relatively few trees were affected by antler damage, but these injuries could be locally significant where tree densities are low. On grouse moors, regular heather burning was principally responsible for the absence of natural woodland. Tree mortality from insects and fungi was negligible. Deaths from climatic agencies, notably drought and frost lift, were fairly few and were largely confined to very small trees. Snowbreak might have locally serious effects on the development of larger trees in winters of exceptional snowfall.

Many species from economically important genera remain rooting recalcitrant, prohibiting the commercialisation of many species in forestry and horticulture, and hindering genetic improvement by conventional breeding or recombinant DNA technology, where vegetative propagation is often used to preserve the genetic fidelity of elite progeny. Two cherry species (Prunus avium and P. padus) were used as models in this study to investigate the physiological and genetic manipulation of adventitious rooting. Mature trees are typically more difficult to propagate vegetatively than their juvenile counterparts. For some trees, micropropagation can circumvent certain effects of ageing and maturation, restoring shoot vigour and rooting, but the mechanism(s) involved have not been elucidated. During micropropagation, subculture interval was found not to be the predominant factor promoting the 'apparent rejuvenation' of mature P. avium tissue. 'Apparently rejuvenated' ex vitro and hedged (putatively) mature P. avium trees were treated with gibberellins predicted to have a range of structural related activities. GA, improved the rooting of cuttings from hedged (putatively) mature cherry, but not from ex vitro trees. Methodology to regenerate adventitious shoots from P. avium leaf explants was developed, (putative) transgenic P. padus plants were produced by an Agrobacterium tumefaciens mediated strategy. Auxin redistribution in planta is postulated to require a component of active transport; inhibition of the predominantly basipetal transport has profound effects on rooting. The putative function of the Arabidopsis thaliana AtAUX1 gene is that of a cellular auxin influx carrier, possibly, as described by the chemiosmotic hypothesis. This thesis examined the hypothesis that transformation with the AtAUX1 gene would enhance the delivery of the root-inducing signal to improve rooting of P. padus, a species which is rooting recalcitrant and more or less obligate on exogenous auxin for this process. However, all six, constitutively expressed, Cauliflower Mosaic Virus 35S promoter driven, 35S::AtAUX1, transgenic shoot lines had reduced rooting.

The establishment and intensive management of forests for the production of timber can have significant effects on the soil carbon dynamics. The establishment of forest on organic soils under grasslands may lead to substantial losses in soil carbon, due to the site preparation for the planting of trees and other disturbances. This is gradually compensated by carbon accumulation in tree biomass as the forest grows until clearfelling at maturity may initiate another substantial carbon loss. This study had two main aims. The first was to investigate the long-term effects of forest establishment on natural grassland as well as clearfelling and re-growth of the forest during second rotation, by looking at the changes in soil carbon stocks and soil carbon balance in a Sitka spruce (Picea sitchsensis) in Harwood (N. E. England). Secondly, to investigate the short-term effects of forest clearfelling on the fluxes of soil CO2, N2O and CH4 and on the environmental factors (soil temperature, water content and water table depth) affecting them. The fluxes were initially measured in two mature stands (40-years old) during one growth season. One of the two stands was subsequently clearfelled while the other was kept intact and fluxes were measured for a further ten months after clearfelling. The relationships between these fluxes and the environmental factors were also examined. The study also investigated the spatial variability of soil CO2 emissions using geostatistical approaches. The soil CO2 fluxes were measured with two methods, a closed dynamic chamber and a closed static chamber, giving the opportunity to compare their relative performance. A performance further investigation on this discrepancy between the two methods took place in lab experiments and on a soil monolith, excavated from the 40-year old stand and kept under controlled conditions in the greenhouse.

Rural Development Forestry (RDF) is being widely promoted in Scotland by both policy makers and pressure groups alike. The idea behind RDF is to recognise and respond to the needs and demands of local communities for forests and forest products. Despite publications and discussion regarding the relative merits of the adoption of an RDF-style of forest management, there is a lack of empirical studies on the subject. This study develops an appraisal technique to investigate forest benefits currently provided by forestry to local communities in Scotland. It examines the future benefits which these same local communities would like to see delivered by the forest and the level of participation in forest management and decision making which they would like, in order for this to be achieved. The appraisal was achieved through surveys, conducted at four different study sites throughout Scotland. The questionnaires utilised during the surveys explored the current benefits which individual respondents derive from their local forests and those delivered to the whole community. In most cases, these were benefits associated with recreational use or amenity values of the forests. For the majority of respondents the most important benefit for the forests to provide in the future is locally based employment. However, the recreational and amenity benefits are also afforded high priority for provision in the future. A contingent valuation survey was utilised to establish the level of demand for participation in forest management and decision-making by rural residents. The results demonstrate that although a number of respondents value the idea of being consulted and having some input into this process, the majority are not in favour of the community taking on responsibility for the forests. However, there is marked variation between the localities. The study highlights a number of implications for future forest policy, if RDF is to be carried forward. One of the most significant findings of the study is the heterogeneity within and between communities and thus the flexible approach which is required for the promotion of RDF.

As the global population continues to increase, so will the demand for timber (and other raw materials) for building, construction, and also for the pulping industry. The high demand for wood and the increasing human population mean that natural forests are being lost and degraded. A potential solution to this problem is to improve the productivity of our plantation forests to relieve the pressure on natural forests in terms of sustainable wood production. This project is the first to use newly available microarray technology to study differential gene expression in cambial tissue of high versus low wood density field grown samples from two contrasting species, Sitka spruce (gymnosperms) and poplar (angiosperms). Genes up-regulated in high-density Sitka spruce and poplar samples had functions in cell formation and expansion, with down-regulated genes having functions in lignin biosynthesis, stress-response and defence. Plantation trees could be screened at a young age to assess their expression of candidate genes to speed up the breeding and selection process.

Kinetic profiles were investigated for the pyridine catalysed reaction of Corsican (CP) and Scots pine (SP) sapwood with a homologous series of linear chain carboxylic anhydrides namely, acetic (AA), propionic (PA), butyric, valeric and hexanoic (HA). With AA. it has been found that the reaction profiles are described by a model where diffusion dominates the reaction process, that is to say that reaction of the reagent molecules with a specific reaction site is rapid compared with diffusion. With longer chain anhydrides, the rate of chemical reaction and diffusion both contributed to the reaction kinetics. The reaction activation energies (Ea) were detennined for the catalysed reaction of pyridine swollen pine sapwood samples and phenolic model compounds (in solution) with the series of anhydrides, using the methods of initial rates and rate constants. Both methods resulted in comparable values. The wood species did not influence the Ea. With wood samples, the Ea was largest for the reaction of AA, and decreased as the molecular weight of the anhydride increased. When the reactions were performed in homogeneous solutions, there was no correlation molecular weight of anhydride and Ea. It is suggested that the lower values obtained for the Ea for reaction with wood are related to the restricted space surrounding the accessible hydroxyl groups (OH). The cell wall micropore network of oven-dried (OD) CP and SP sapwood was investigated by reaction with AA and PA in a non-swelling solvent (xylene) and with swollen wood. Significant differences in reactivity were found between species. The volumetric changes in CP and SP sapwood due to modification with the series of anhydrides were studied. Again significant differences in the response of the wood cell wall were found between species. A comprehensive investigation into the water sorptive properties and into the effect of molecular size of the substituent group upon the sorption of water vapour of softwood modified with the series of anhydrides was perfonned. The sorption isotherms for untreated and chemically modified wood were analysed using the Hailwood-Horrobin model. The results are interpreted by consideration of both the number of OH groups reacted and the volume occupied by adduct in the cell wall. It is considered that the latter effect is more important. In the final part of this study, an approach was made to investigate the cell wall porosity of unmodified and modified wood, using the nitrogen adsorption technique. Results indicated that the true porosity of the cell wall was not determined by this technique.