School of Agriculture, Food and Ecosystem Sciences - Theses

Permanent URI for this collection

http://hdl.handle.net/11343/427

Search Results

Now showing 1 - 10 of 138

Soil water and nitrogen dynamics of dryland wheat in the Victorian Wimmera and Mallee

O'Leary, G. J. (Garry John), 1953- (University of Melbourne, 1994)

In semi-arid environments where water supply is the major factor limiting grain yield, the agronomic option of fallowing provides a way to increase water and nitrogen (N) supply. A study was carried out for four years at two sites in semiarid, north-west Victoria, Australia to examine the effect of fallow management on water and N supply at sowing and their subsequent effect on growth and yield of wheat. The fallow treatments comprised the four combinations of stubble management (retention - SR, removal - NS) and tillage (frequent - CT, none - NT) in year-long fallows of a fallow-wheat rotation (i.e. SRNT, SRCT, NSNT and NSCT), together with a tilled summer fallow prior to alternating pea and wheat crops (PeCT). The two sites, which have different rainfall, represent the major soil types of the region; a chromic vertisol at Dooen and a calcic xerosol at Walpeup. At Dooen, grain yield provided a diverse response to fallow method that generally reflected biomass response and water use. NSCT increased yield every year over PeCT by an average of 0.9 t ha-1 (range 0.5 to 1.3 t ha-1). Stubble retention (SRNT and SRCT) further increased yield above the non-stubble fallows (NSNT and NSCT) in three of the four years, averaging an additional 0.56 t ha-1 (range; 0.24 to 1.04 t ha-1). At Walpeup, NSCT increased yield above PeCT in two years, by 0.7 t ha-1 from 2.6 to 3.2 t ha-1 in 1990 and by 0.6 t ha-1 from 1.7 to 2.3 t ha-1 in 1991. Stubble retention increased yield by 0.45 t ha-1 in only one year, 1988, and that was without tillage (SRNT). Zero tillage did not have any independent overall effect on yield. Crops responded differently to the distinct soil conditions at each site. At Dooen, NSCT increased soil water storage at sowing by an average of 76 mm (range 24-122 mm) above PeCT. Stubble retention added a further 52 mm (range 36-65 mm). Zero tillage was slightly beneficial, and was enhanced by the presence of stubble. Similarly, at Walpeup, NSCT provided an average of 37 mm (range 23-56 mm) more water at sowing than PeCT. SRNT increased water storage by 27 mm in only one year. Fallow and its surface treatments provided large differences in soil mineral nitrogen (SMN) accumulation, but levels were generally sufficiently high (66-222 kg N ha-1) to not greatly affect yield. At Dooen, NSCT accumulated 46 kg N ha-1 more SMN than PeCT, whereas at Walpeup PeCT increased SMN in two years by an average of 46 kg N ha-1 above NSCT. At Dooen stubble retention depressed SMN by an average of 26 kg N ha-1 in two years and in one, increased it by 20 kg N ha-1. Stubble retention had little effect on SMN at Walpeup. A simulation model was developed to apply the responses of the field experiment to a wider range of weather conditions. It includes soil carbon (C) and N and crop N submodels. The model showed a pattern consistent with observed response. Validation showed that soil water was simulated more accurately than SMN (R2=0.95 cf. R2=0.61), and there was no significant bias in any fallow treatment. Crop biomass followed the general pattern of observations throughout the season well (R2 =0.77). Grain yield was simulated with less accuracy than biomass (R2 =0.42) and was shown to be largely due to poor simulation of grain number (R2 =0.07) because accuracy of yield prediction was markedly increased (R2 =0.59) when predictions were made on observed grain numbers. Grain [N] was simulated with fair accuracy (R2 =0.29), pointing to the need for improvements not only in the simulation of grain size (R2=0.61), but also in the partitioning of N to grain. Overall, the performance of the model is comparable to the CERFS-Wheat-N model even though it is less complex. Analyses with the model over long-term weather conditions supported the field observations of the advantage of stubble retention in predicting a median increase in yield of 0.8 and 0.6 t ha-1 at Dooen and Walpeup, respectively. However, at Walpeup, field observations indicate that other factors may limit this advantage. Zero tillage had a small advantage in the presence of stubble (0.2 t ha-1) with no consistent advantage without stubble at Dooen nor under any system at Walpeup. The model displays a general level of behaviour known to occur in the field and provides a useful tool, not previously available, to examine water and N interactions in conservation tillage systems in north-west Victoria. The model should have relevance in other semi-arid environments where wheat is grown.
Villager participation in native forest development : the impact of social forestry in Fiji

Vize, Susan Maree (University of Melbourne, 1993)

As owners of extensive forest resources, the Fijian people have an opportunity to utilise these resources for village development. Villagers are very conscious of their low living standards compared with urban areas. It is widely accepted that expanding local employment and business opportunities is necessary to improve rural living standards, and rural Fijians hope that the forest industry will help them to achieve this. The object of this research project is to evaluate the success of social forestry as a methodology for achieving village development in Fiji, particularly the contribution of Fijian villager participation in the native forest industry. Forestry activities earn cash benefits for rural villagers in the forms of land rentals and timber royalties. Social forestry programmes involving villager participation in planting projects were seen as way to enhance rural development. To further increase the benefits obtained from forestry many Fijians have become actively involved in forest harvesting and timber processing. Though there is a high potential for achieving significant benefits by Fijian participation in the forestry industry, the actual benefits received are determined by the type of activity, the level of participation and the performance of the participants. The study involved surveying in seven villages on the islands of Viti Levu and Vanua Levu, as well as short interviews with many Fijians involved in the forest industry. The villages covered a range of villager forestry participation types, including casual and contract employment in logging or planting, village community fund-raising contracts, a proposed joint venture logging project and a small logging company. 'Development' appears to be fairly consistently interpreted across rural Fiji as the improvement of village services and homes; hence in the eyes of the villagers, forestry does contribute to development. The surveys show that forestry can make a significant contribution to the village economy in the form of cash and employment, but the contribution of forestry towards village development is dependent upon the use of the income generated from forestry. These uses vary between the villages, but commonly include the purchase of consumables, such as food and tobacco, paying school fees, home improvements, roads, churches, schools, and other communal amenities. There is a significant difference in the contribution of forestry to employment and income between the two islands; forestry is more important in rural development on Vanua Levu. Social structure in the village, village leadership, assistance from government or financial advisers, and the level of integration of the village into the market economy, all appear to affect the investment and management of forestry funds. These aspects exert the most influence on the amount of 'development' that can be achieved through forestry programmes involving Fijian landowners. Logging businesses achieved higher cash returns than other types of Fijian forestry participation, but were plagued by difficulties. Businesses in logging, sawmilling and manufacturing of higher value-added products such as veneer and furniture are the most profitable area of the forest industries in Fiji. The potential for Fijian villagers to become involved in these enterprises is severely limited by access to credit as well as their paucity of management and administrative skills. But where the potential for forestry participation exists and motivation is provided by a community leader or individual, social forestry activities can effectively promote rural development and viable village communities in Fiji.
Estimation of timber volumes in eucalypt forest using large-scale aerial photography and ground samples

Biggs, Paul Haydn (University of Melbourne, 1991)

This study investigated the use of large-scale aerial photographs (1:1200) for estimating timber volumes in eucalypt forest in Western Australia. The aims were to determine the optimum photography system for estimating timber volumes in jarrah forest, to derive appropriate aerial volume functions, and to investigate the efficiency of different sampling designs using aerial photographs. Photographs were acquired using a fixed-base photographic system, the cameras being mounted on a transverse boom on a helicopter. Two types of cameras, Vinten and Hasselblad, were tested. Measurement of the Vinten photographs using simple stereometers provided estimates of tree heights with a standard deviation of 1.7 m, while the standard deviation of measurements on Hasselblad photographs was about 2.0 m. These results are slightly poorer than results achieved by others in coniferous forests, the reason being attributed to imprecision of the ground "truth" rather than poor quality in the photographs. The accuracy of species interpretation was 85 percent on all photographs. It is expected that more precise measurements could be obtained using analytical plotters to measure Hasselblad photographs. Hasselblad 500 EL/M cameras used in the present study were fitted with reseau plates and were calibrated to facilitate analytical measurement. Results showed, however, that further research is required into the process of exterior orientation with such large-scale photographs with a very small air-base/flying height ratio. A system of navigation and camera-control was built specially for this project using a Global Positioning System (GPS) receiver and a laptop computer. This system was used to navigate during photography and to determine the location of each pair of photographs on maps and in the field. Its use led to efficient in-flight navigation and ensured that photographs were taken on a fixed-distance interval. Ground plots could be located quickly in the forest using this system, as the actual photo-centres were generally within 50 m of their surveyed position based on GPS coordinates. Sample data from both photo and ground plots were stored in a Geographic Information System (GIS) database, which allowed them to be linked with other spatial data. Using this system, volume statements could be derived for any part of the forest by defining the relevant geographical zone and then identifying the plots within that zone. Any change in the available land base could also be accomodated, simply by redefining the zone of interest and reselecting the sample plots. Aerial tree volume functions derived for different species in the jarrah forest used the total height of each tree and the dominant height of the stand as independent variables. Separate functions were derived for different areas of forest by applying a nominal dominant height value in a general volume function, while another set was derived by stratifying the original sample plots. It was found that the volumes of very large trees were greatly underestimated when their crowns had been damaged, and that this factor swamped the difference between the two types of volume functions. The underestimates of photo volume caused by the damaged crowns of large trees had the effect of reducing the correlation between the photo and ground samples to about r=0.65. The cost of each photo plot was about $30 and the cost of each ground plot was $280. The combination of these costs and correlation meant that double sampling for regression could provide estimates of volume at the same cost and precision as ground sampling. Further research should, however, investigate ways to improve this correlation to about r=0.75, which would then make double sampling with large-scale photographs cheaper than ground sampling.
Innovation in response to environmental problems

Ewers, Craig. (University of Melbourne, 1990)

An important and long-standing field of agricultural extension research has involved the acceptance of new agricultural practices. Consistent with the applied, purposive nature of extension research and the emphasis on agricultural development, this field has centred around the question: How do we get farmers to adopt new technology? As such, the process has traditionally been viewed as one which is confined to the decisions of individual farmers. The role of those who develop and distribute the technology has generally been overlooked. It is argued in this thesis that when the development and promotion of agricultural technology are an integral part of government strategy, a great deal of decision-making happens before the technology even reaches the farm. Indeed, it will be shown that some innovations developed by government scientists may not become fully available to farmers. The complex interplay between science, technology and government policy forms a central theme throughout this thesis. Changes in agricultural practice are discussed with reference to the social and economic conditions that triggered them. In particular, the growing impact of the environmental movement on the development and diffusion of new cropping techniques throughout the 1980s is covered in depth. Two current Victorian government programs are critically reviewed along these lines. Finally, the views of farmers are analysed, through a survey of 146 wheat growers in North-East Victoria.
A comparison of cattle management systems in Kupang district, East Nusa Tenggara province, Indonesia

Kapa, Maximilian Marthen Josephus ( 1994)

A program of fieldwork was conducted in Kupang District, East Nusa Tenggara Province, Indonesia for a period of three months. The aim of this study was to compare two systems of cattle rearing: the Paron (intensive) and the non-Paron (extensive) systems. Six farmers from extensive and 38 farmers for paron systems were chosen as respondents. In addition, selected characteristics were measured of the 101 cattle from the paron (36 females and 65 males) and 199 cattle from the extensive systems (134 females and 65 males). The study showed that one hundred per cent of respondents engaged in agriculture enterprises. Mixed cropping was practised and the main food crops cultivated were maize, cassava, and beans. Livestock enterprises, mainly beef cattle, were an integral part of these farming systems. This study sought to establish the growth rates for cattle within these particular systems. The data collected gave strong correlations between girth and weight, and fairly strong correlations between weight and age, weight and value, and age and value. Under village conditions growth rates are quite slow, characterised by relatively fast growth in the early stages (0-9 months) and reducing thereafter (> 12 months). Both male and female paron cattle grew more quickly than those on the extensive system. A major significant input factor determining profitability is labour and, particularly, labour used to feed the cattle. The paron system required more labour to supply fodders but fattened animals for two years (from 200 to 300 kg), while the extensive system required at least five years to reach the preferred weight (150 to 300 kg). Financial analyses used Gross Margin analysis (Cash Gross Margin and Imputed Gross Margin) as proxies for profitability. Various constraints deter the productivity of the two systems: lack of fodders and water availability resulting from unpredictable rainfall, as well as total availability of labour. It is suggested that there is sufficient way available in the paron system to boost profitability, through increasing growth rate. Currently, this averages some 180 g per day over the fattening period. Improving this through better feeding management will increase profitability, depending upon the additional labour required. A series of limits to these changes is presented.
Sclerotinia minor in sunflowers : onset of disease and bioprotection with gliocladium virens

Burgess, Diana R ( 1994)

Sclerotinia stem rot (Sclerotinia minor) is a serious constraint to sunflower production in south-eastern Australia, where the lack of cultivars with adequate levels of resistance and the costs of chemical control make biocontrol an attractive option. Onset of disease in the field occurs at bud development. Field trials with the highly susceptible inbred line, PacF2582, confirmed the absence of disease during vegetative growth. A study of root development found a marked proliferation of lateral roots at GS 3.1-3.3 (early to mid-bud) in field-grown sunflowers and slightly earlier in glasshouse plants. Glasshouse studies demonstrated the susceptibility of vegetative stage plants to infection of lateral roots by one or 2 pre-germinated sclerotia. Onset of disease in soil inoculated to a high density with sclerotia was also delayed, generally until GS 1.5 (8 leaf), suggesting that occasional contact between roots and sclerotia is not sufficient for disease initiation. A model system was developed to examine the effects of surface microflora and exogenous nutrient on sclerotial germination. Field sclerotia, incubated on water agar disks, required surface sterilisation to relieve fungistatic dormancy. Root sap and exudates from sunflower lines stimulated germination of partially surface sterilised field sclerotia, as did root sap from non-host species. Germination rates were significantly higher with root nutrient from plants during vegetative growth than from plants at mid-bud stage. Exposure to root sap for more than 3 days was required for stimulation of germination in vitro. The germination response of soilconditioned culture sclerotia to nutrients was less consistent than that of field sclerotia. The results indicate that root nutrients have potential to stimulate germination during vegetative growth, but onset of disease awaits root proliferation at bud development. It was proposed that a threshold of root activity prolongs the exposure of sclerotia to root exudates, while overlapping rhizospheres may provide a critical concentration of nutrient for germination and mycelial development. An isolate of Gliocladium virens from disease-affected soil proved highly antagonistic to S. minor in culture. Seed treatment with G. virens suppressed disease induced by inoculation of lateral roots in pasteurised potting medium and in field soil. Significant disease control was also obtained by sowing seed into a plug of G. virens growing on V8-vermiculite or by pre-germination of seed with a culture of G. virens. Bioprotection by seed treatment with G. virens was evaluated at two field sites, and was effective in field microplots where disease incidence was less than 50%. Since G. virens, applied on seed, did not affect root growth and did not grow into the lateral root zone, it was concluded that seed treatment with G. virens creates a "cordon sanitaire" around the upper tap root, sufficient for bioprotection of inoculated plants in pots and field microplots. If seed priming techniques can be adapted to allow germination of G. virens on seed before sowing, it may be possible to enhance the cordon sanitaire and delay onset of disease in the field to obtain commercial yields.
A simulation study of nitrogen fertilizer management strategies for wheat crops in North-West Victoria

Bheenick, Krishan Jeehan ( 1994)

This study addresses the issue of nitrogen fertilizer application in wheat crops under variable climate in the dryland Mallee and Wimmera districts of north-west Victoria. A crop simulation model is used to determine the response of crops to various management strategies over a number of years, using generated weather data characteristic of the study area. In the first stage of the study, the performance of a weather generator is tested. for accurate representation of weather data at a range of sites so that it can then be used to provide daily weather data for input to crop models. A weather. generator which requires only monthly weather statistics as input was tested (SIMMETEO). It overpredicted the annual number of wet days and rainfall by less than 6% of annual rainfall, which would cause small errors in crop simulations results. Generated and observed mean monthly temperatures were not significantly different. Mean monthly solar radiation was underestimated by up to 2.8 MJm-2day-1. This was found to be mostly related to the prediction of solar radiation on wet days and a solution has been suggested for sites in southern Australia. A method for the estimation of mean, monthly solar radiation from latitude was successfully tested and this can be used to provide input data for the weather generator at sites where long-term data for solar radiation are not available. The weather generator was found suitable, for use with crop simulation models in the temperate wheat-producing areas of the Australian wheat belt, where it performed better than in the tropics. In the second stage of the study the performance of a crop simulation model was tested in north-west Victoria. The CERES-Wheat-N model was tested using data gathered from commercial crops with various nitrogen fertilizer treatments from two farms over two gears at Donald and Warracknabeal. Crop growth was overestimated early in the season, which also caused the overestimation of nitrogen uptake, especially in the fertilized treatments. Soil moisture was often underestimated towards the end of the season. Nevertheless, grain yields (range; 0.6-3.7 t ha-1) were better predicted (r2:0.79) than these crop growth components and reflected known crop responses to fertilizer application and rainfall variation. The CERES-Wheat-N model requires further testing in dryland agriculture in Victoria but it can presently provide valuable information for the management of wheat crops. The performance of the model in this test is discussed in relation to other studies reported over a wider range of sites and environments. In the final stage of the study SIMMETEO and CERES-Wheat-N were.used.to examine the response of wheat crops in north-west Victoria to nitrogen fertilizer application under variable rainfall, associated with El Nino-Southern Oscillation (ENSO) and anti-ENSO episodes. These simulations were run for three sites in the Wimmera and Mallee, namely Donald, Horsham and Ouyen. Wheat production was simulated for a range of combinations of soil type, sowing date and fertilizer application over 50 years of generated weather data typical of overall, ENSO and anti-ENSO years. Simulated grain yields ranged from 0.2-3.8 t ha-1 at Ouyen, 0.3-3.8 t ha-1 at Donald and 2.3-3.8 t ha-1 at Horsham, without any applied fertilizer. The results for each combination were summarised, as exceedence probability curves for yield. These provide information about the probability of yield exceeding a certain level over the climatic range at the site. The results indicated that wheat variety Matong produced die highest yields when sown in May compared to June and July, given that frost damage at flowering is not considered. The highest yields occurred on the heavy clay soil at Horsham, followed by the grey-clay loam at Donald and the sandy loam at Ouyen. Crops on these soils also showed varied response to nitrogen application, mostly dependent on soil moisture during the season. Crop yields were higher in anti-ENSO years and lower in ENSO years, compared to overall years. Yield response to rainfall variability was marked on the sandy soil and the clay loam but not on the heavy, clay which had sufficient soil moisture to support crop growth, even in ENSO years. Applied nitrogen caused yield reductions in some ENSO years, sometimes seen as a cross-over of exceedence probability curves. The presentation of the simulation results as exceedence probability curves provides information to help decide how much nitrogen can be safely applied to achieve a particular yield with a specified risk of crop failure. Thus, wheat growers are able to identify a fertilizer management strategy that will give the most benefit in the long run, depending on the economics of production on individual farms and their attitude to risk. This study has demonstrated the potential value of using crop growth models and weather generators in crop management and has also contributed towards identifying areas in which research is required to improve the potential for the practical application of these simulation techniques.
An economic analysis of Nigeria's import substitution policy for rice and wheat

Matane, Ibrahim Ahmed ( 1993)

The study involves analyses of substitution effects of Nigeria's import restriction imposed on rice and wheat; in terms of domestic production, consumption and prices. The theoretical and empirical literature on uncertainty, risk and self sufficiency under international trade is reviewed, as well as smuggling and its welfare implications. The analyses involves use of regression and correlation models. The result shows that increase in price of wheat would reduce per capita wheat consumption, thereby shifting consumption to maize and millet. It is found that wheat and rice are inferior commodities in Nigeria. In addition, increase in the prices of yam and maize would increase per capita rice consumption . Therefore, yam, maize and rice are substitutes in consumption, so also are wheat and rice in production. An increase in price of rice would increase rice area while increases in prices of maize and wheat would increase wheat area. While increase in the price of rice would lower wheat production, thereby favouring rice production in the longrun. Accordingly, increase in world prices for rice and wheat would lower their domestic prices. While increase in tariff would increase domestic wheat producer price. Between 1973 and 1986 prices of most food commodities in Nigeria increased both nominally and in real terms. In addition, real prices for rice and wheat slumped between 1987 and 1989. Retail prices were always higher than producer prices. It is concluded that insulating domestic prices and supplies of rice and wheat is no panacea to problems associated with price instability, self sufficiency and balance of payments. Sustainable policies, such as consumption tax; use of futures markets; diversification and appropriate exchange rate are the best policy options.
Pulpwood testing and evaluation : a case study for a proposed kraft pulp mill and associated plantations in Zimbabwe

Muneri, A ( 1994)

Acacia mearnsii, Eucalyptus grandis and Pinus patula are the main commercial plantation species grown in the Eastern Highlands of Zimbabwe. They are a potential source of pulpwood for a proposed bleached kraft pulp and paper mill. The aim of this study was to investigate and evaluate the technical factors to be considered when planning pulpwood plantations. An extensive literature survey of important pulpwood properties, with emphasis on measurable and potentially controllable properties, revealed that neither cellulose, lignin nor extractive content are universally reliable indicators of pulp yield. Wood density emerged as the most important pulpwood property affecting wood handling costs, pulp yield, pulp quality and digester productivity. Among the fibre properties, fibre length and fibre coarseness were found to have large influence on paper properties although this is more important where large differences occur such as those found between softwoods and hardwoods. To provide material for assessing the pulpwood quality, the hardwood species were sampled at single sites from 8 year old trees, while P. patula was sampled at two sites (Erin and Stapleford) from two age classes, 14 and 25 years (ie. 6 stands). For investigation of wood properties within trees, 25 mm thick discs were taken at six height levels (0, 10; 30, 50, 70 and 90% up to 8 cm diameter over bark) from three sample trees from each of the six stands. The material was also used for pulping within tree of P. patula, by simulating utilisation of 7 year old thinnings, toplogs and slabwood. In order to determine the average pulpwood quality, six trees were sampled from each of the P. patula stands and twelve trees from each of the hardwood stands. A single log (15 cm long) was taken from the trees in rotation from the midpoint of the butt, middle and top third of the stem. Wood density was measured by an x-ray densitometric technique along the north and south radii of all sample trees. Fibre length and coarseness were measured by the Kajaani FS200 fibre analyser. Investigation of fibre length was carried out on the north radii of a single tree from each of the four P. patula site-age classes, and two trees of each of the hardwoods. Fibre coarseness was only investigated on a single tree of each hardwood species. P. patula fibres caused blockages of the Kajaani capillary which frequently stopped the analysis and invalidated the measurement. A technique was developed for preparing whole fibres for determining length distributions and coarseness. Radial ships, while clamped between perforated metal plates, were partially delignified by kraft pulping and bundles of whole fibres were extracted from the pulped still), thereby avoiding the problem posed by cut and broken fibres. A special procedure was developed, using Millipore filters, to estimate oven-dry mass of the fibres in the suspension used in the Kajaani fibre analyser and through this the fibre coarseness. Wood properties were measured and recorded for each growth ring for P. patula and for every 5 mm section from the pith for the hardwoods. Within-tree data for P. patula were analysed on the basis of the age of tree when the growth ring was formed (year-number). Analyses of variance were used ito examine differences in density, fibre length and fibre coarseness between trees, between the north and south radii, between six height levels and between year- numbers or distances from pith. Regression analysis was used to estimate mathematical equations that best described the pattern of the wood property within the tree. Predicted values calculated from the equations were presented as grey scale maps to illustrate visually the within-tree variation pattern of the properties. P. patula showed an increase in wood density with tree age. The area-weighted mean densities of the 14 year old trees of P. patula were 0.476 and 0.492 g/cm3 respectively for samples from Erin and Stapleford. The respective values for the 25 year old trees were 0.557 and 0.544g/cm3. A. mearnsii and E. grandis had area-weighted mean density values of 0.811 and 0.583 g/cm3 respectively. For the three species, variation of density between trees, between heights and with year-number or distance from pith was statistically significant. For P. patula and A. mearnsii, density decreased with tree height, while density increased after an initial decrease from 0 to 10% height level in E. grandis. In both P. patula and A. mearnsii, the overall decrease between 0% and 90% height levels was about 20% for all trees. At all height levels, density increased from pith to bark. For P. patula the rate of increase from pith bark was similar at the different heights, resulting in a cylindrically symmetric within-tree pattern, similar to that reported for P. radiata (Cown & McConchie 1980). A similar pattern was found in A. mearnsii, but the central core of low density extended from about 30% height level upwards. The pattern of wood density in E. grandis requires further work since the regression equation obtained had a low coefficient of determination and there were no obvious trends apparent. Equations from previous studies (e.g. Malan 1988) also failed to describe the density pattern. The corresponding points along the two radii of both P. patula and A. mearnsii were found not to differ significantly. Although a significant difference was observed in E. grandis, the difference was small in practical terms. Fibre length in P. patula increased with tree age. The ring area-weighted mean fibre length for the 14 year old trees were 3.48 mm and 3.07 mm respectively for the samples from Erin and Stapleford. Fibre lengths for the 25 year old trees were 3.55 and 3.84 mm respectively for Erin and Stapleford. The fibre length of 0.89 mm for E. grandis was significantly (at the 5% probability level) longer than the 0.76 mm recorded for A. mearnsii. Fibre length varied significantly with tree height and with year-number or distance from pith. The general pattern in all three species was for fibre length to increase from a minimum at the butt, reaching a maximum at 10 or 30% height level before decreasing towards the apex. Radial increase from pith to bark was more than 100% at most height levels. In P. patula, most of this increase occurred in the first 10 growth rings from the pith. Unlike wood density, the within tree pattern for fibre length was nearly symmetric conically. These patterns closely resemble those reported for P. radiata (Cown & McConchie 1980). Both hardwood species exhibited a cylindrical distribution of fibre length within-tree. This difference arises from the linear radial increase found in the hardwoods as opposed to the curvilinear increase found in P.patula. E. grandis had coarser fibres, yielding an area-weighted fibre coarseness of 6.19 mg/100M compared to 4.23 mg/100m for A. mearnsit. The trend in the mean fibre coarseness is different in the two species. For A. mearnsu, fibre coarseness increased up the tree but in E. grandis there was a decrease with tree height. In E. grandis, fibre coarseness increased with distance from pith, while in A. mearnsu it remained relatively constant. The interesting point is that the radial trends are the reverse of those found for density. These results contradict the generally held view that density and fibre coarseness are correlated. More research is needed to clarify this because, while the relationship was statistically significant for E. grandis, that for A. mearnsii was not significant. Basic density of the pulpwood samples was 414 kg/m3 for E. grandis, 610 kg/m3 for A. mearnsii, and averaged 370 and 420 kg/m3 for the 14 and 25 year-old P. patula. To achieve a Kappa number of between 20 and 30, P. patula, required 18 to 19% active alkali to produce 44 to 45% screened pulp yield. For Kappa number 20, A. mearnsii required 12.5% active alkali to yield 53.0% screened pulp and E. grandis 10.5% active alkali for 54.2% screened pulp yield. The study demonstrated that the two hardwood species can successfully pulped together resulting in a marginally lower screened yield. A. mearnsii had the highest pulp yield per cubic metre of wood (323 kg) followed by E. grandis (224 kg) and P. patula (174 kg). All pulps could be bleached to a brightness of 87-89% ISO. Lightly beaten pulp from the 25 year old P. patula trees had the highest tear index (18 - 21 mN.m2/g), and highly beaten pulps from the 14 year old P. patula trees had the highest tensile index (112 - 115 N.m/g) and burst index (about 9 kPa.m2/g). Bleaching reduced tear index but increased tensile index. At 300 Csf, the tensile and tear indices for E. grandis pulp (127 N.m/g and 11.0 mN.m2/g) and for A. mearnsii pulp (103 N.m/g and 9.4 mN.m2/g) would be classified as very good and adequate for use in unbleached packaging papers. A. mearnsii pulp had higher opacity (70 vs 65%) which is crucial for printing papers. Bleaching reduced tensile index by about 14% in both species. The pulping and papermaking results of A. mearnsii demonstrates that the conventional prejudice against the use of this species as pulpwood in Zimbabwe may be wrong. Its high density means that on a volume basis, there is more wood material that can be transported and higher pulpwood productivity will enable pulp production from less volume of wood for a given digester capacity. These all have beneficial economic consequences to the pulp producer. The species has the added advantage that it can be successfully co-pulped with the other major hardwood species. A new sampling scheme was developed to provide wood formed when trees were 7 years old. This was useful in demonstrating that pulp yield and papermaking properties of the simulated 7 year old thinnings of P. patula were not influenced by tree age from which they were extracted. Bulk density decreased in the order of slabwood, toplog and 7 year old. Slabwood produced pulp with the highest tear index. This property was about the same for toplogs and 7 year old material. The test results for the P. patula components provided an opportunity for correlation analysis of wood density and fibre length with pulp properties (freeness, bulk, tear index, tensile index, stretch, burst index and air permeance). All handsheet properties examined (except freeness and stretch from high tearing strength pulp) could be predicted in terms of wood density and fibre length to the extent of accounting for at least 69% of the variation. A model was developed for estimating plantation areas required for the production of the proposed pulp grades. It involves three parameters and the associated coefficients of variation, namely pulp yield (Y), basic density of wood (BD) and mean annual volume increment (MA1). Using the growth rates of similar species in South Africa and test results from this study, it is estimated that the 60 000 t/y mill could be supplied from about 20 000 ha of dedicated pulpwood plantations. The results show that growing P. patula pulpwood on 25 year rotation as against 14 year rotation will provide high strength pulps for packaging papers.
Competition between pinus radiata D. Don. and pasture for water

Hadriyanto, Deddy ( 1994)

School of Agriculture, Food and Ecosystem Sciences - Theses

Permanent URI for this collection

Filters

Date

Author

Subject

Type

Settings

Sort By

Results per page

Statistics

Citations

Search Results