No description available.

No description available.

No description available.

Sustainable forest management is a key issue in Malawi. Malawi is a relatively small, resource poor, densely populated country, which in some areas is close to exceeding the energy capacity of the environment to support it. Despite the importance of forestry in Malawi, there is a severe lack of knowledge about the current state of Malawi’s forest resources. Remote sensing has the potential to provide current and historical insights into forest cover change. However, Malawi faces a number of key challenges with regards to in-country remote sensing. These include technical capacity for obtaining accurate and consistent forest area and biomass estimates, with errors at acceptable levels, as well as the necessary supporting capacity development for individuals and institutions. This thesis examines how remote sensing can be used to support sustainable forestry in Malawi, by assessing the use of both optical and Synthetic Aperture Radar (SAR) data for mapping forest cover, forest cover change and aboveground biomass (AGB). L-band SAR data was used to try and establish a relationship between radar backscatter and biomass, which has been achieved many times in other areas. However, no correlations between any field-based forest metric and backscatter explained enough of the variability in the datasets to be used to develop empirical relationships between the variables. There were also differences between my field measured AGB and AGB values predicted by a published backscatter-biomass relationship for African dry forests. The speckle inherent in SAR imagery, the heterogeneity of Malawi’s dominant miombo savanna, and Malawi’s variable topography are likely to have played a significant role in this. Two different MODIS products were investigated for their potential for mapping forest cover change, with regards to potential REDD+ schemes. As part of this, a published equation was used to calculate the break-even point for REDD+ schemes in Malawi, using estimates of forest area and deforestation for the United Nations Forest Resources Assessment 2010. The results of this equation show that measurement error is the most important factor in determining whether or not Malawi can make REDD+ economically viable, particularly at lower levels of deforestation. While neither of the MODIS products were able to produce a verifiable forest cover change map, they do confirm that Malawi is experiencing some level of forest loss, and help to narrow down the range of possible forest loss rates Malawi is experiencing to between 1-3% net forest loss per year. Finally, this thesis examines global trends in the engagement of developing country researchers with global academic remote sensing research, to investigate differences in in-country capacity for monitoring forests using remote sensing. The results of this found that while a significant proportion of Earth observation research (44%) has developing countries as their object of research, less than 3% of publications have authors working, or affiliated to, a developing country (excluding China, India and Brazil, which are not only countries in transition, but have well established EO capacity). These patterns appear consistent over the past 20 years, despite the increasing awareness of the importance of capacity development over this period. Despite inconclusive results from the approaches examined here, remote sensing can play a role in improving understanding about the dynamics of Malawi’s forest resources. There is a need for nationwide accurate, validated forest maps that can be repeated at least on a yearly basis, and remote sensing could produced these without the resources needed to conduct full national ground inventories each year. If remote sensing is to be useful as a forest mapping tool in Malawi, it needs to provide consistent, verifiable and updatable estimates of forest cover and biomass change. This ideally needs to be achieved using free or low cost data, and by using open source or open access software, as this will better enable incountry researchers to conduct on-going forest mapping activities.

This thesis investigates the influence of niche specialisation on floristic composition and species distributions of the family Dipterocarpaceae along topographic gradients at two contrasting lowland Mixed Dipterocarp Forest sites in Brunei Darussalam: Andulau (overlying nutrient-poor sandy soils) and Belalong (overlying nutrient-rich clay soils). Dipterocarp tree density, species richness and diversity were higher at Andulau than Belalong, and were also higher on ridges than in slopes and valleys. Randomisation tests detected significant dipterocarp species associations with site and tomography. Dipterocarp floristic composition was strongly correlated with various habitat variables at Belalong, even after accounting for distance between samples. In contrast, fewer habitat variables were correlated with dipterocarp floristic composition at Andulau, implying weaker habitat effects in this more topographically homogeneous site. In a field-based transplant experiment, <i>Dryobalanops aromatica </i>and <i>Dryobalanops rappa </i>seedlings consistently showed higher growth rates and survival in gaps than in understorey plots. Higher survival and leaf production of <i>D. aromatica </i>seedlings at Andulau than Belalong provide evidence of a habitat preference. Growth and survival of <i>D. rappa</i> seedlings were similar at the two sites, but <i>D. rappa</i> seedlings grew significantly faster in height than <i>D. aromatica </i>seedlings on slopes at both sites. In conclusion, local and landscape scale variations in edaphic and environmental resources influence dipterocarp species distributions and floristic composition, as well as dipterocarp seedling growth and survival. Thus, niche specialisation is an important mechanism in the maintenance of species coexistence at Andulau and Belalong.

Student Number : 9906840R - MSc research report - School of Animal, Plant and Environmental Science / The global increased demand for forest products has led to an increase in the area of exotic fast-growing forest plantations. An understanding of nutrient cycling in plantations is essential to enhance their productivity. Sustainable forest productivity involves the managing of nutrients and genetic factors to maximize yields such that they are increasing or non-declining through the maintenance of soil quality and selection of superior tree species and breeds. Komatiland Forests Limited (KFL), a South African forestry company, initiated a permanent sampling plot (PSP) programme in 1998, where it monitors over 30 foliar and soil parameters, as well as tree growth parameters. This study utilized a subset of the permanent sample plots (PSPs) database to compile a suite of foliar and soil parameters that can be used to better interpret stand productivity in pine plantations. Data from PSPs of pine species Pinus patula, Pinus elliottii and the hybrid Pinus elliottii x caribaea planted on dolomite, granite and shale were used in the statistical analyses as they were well represented in the dataset. The geological analysis revealed that parent material significantly affects soil organic carbon content; soil exchangeable K, soil Fe, soil Mn and foliar Mn concentrations. Exchangeable K was found to be low across the geologies ranging on average from 0.08 – 0.11cmol/kg. An accumulative effect was found in foliar concentration of Mn across the geologies and species, with average foliar Mn concentrations being as high as 1086ppm. No statistically significant differences were found at the geological level in soil N, P, exchangeable Ca, Mg & Na, Al, pH or soil texture. Neither were there any significant differences in foliar concentration of N, P, Ca, Mg, Na, Cu, Fe, Zn, B and S at the geological level of analysis. However significant correlations were found between soil cation exchange capacity, soil pH and foliar concentration of Zn, Mn, Mg and Ca. Pinus patula had significantly higher foliar concentrations of N (p<0.001), P (p<0.001). Mg (p=0.001), B (p=0.001) and S (p<0.001) than the other pine variants under analysis. However when species x geology interaction analysis was used P. patula only had significantly higher foliar concentrations with regard to N (p<0.001) and P (p<0.001), and lower foliar concentrations of Zn (p<0.001) and Na (p=0.041) than the other pine variants under analysis. Across the species and geologies, soil acidification resulted in low Ca (0.15-1.6 cmol/kg) and Mg (0.1-0.7cmol/kg) availability. Positive and significant correlations were found between foliar and soil concentrations of N (p=0.022), P (p=0.030), Mg (p<0.001) and Ca (p<0.001). Productivity of the hybrid was significantly higher than the other two species (p<0.001), while P. elliottii had significantly lower productivity than P. patula (p=0.001). Regression models and a principal component analyses revealed that from the dataset of soil and foliar chemical and physical parameters Mg Soil, CEC, N soil, N foliar, P foliar, K foliar, Cu foliar, B foliar, S foliar, C:N soil, Ca:Al soil, N:Ca foliar, N:K foliar, clay and silt are best correlated with stand productivity.

1v. (various paging) : / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (D.Sc.1979)--University of Adelaide, Dept. of Physical and Inorganic Chemistry, 1979

Thesis (M.S.)--Oregon State University, 1965. / Typescript. Includes bibliographical references (leaves 34-35). Also available on the World Wide Web.

Thesis (Ph. D.)--Oregon State University, 1971. / Typescript (photocopy). Includes bibliographical references. Also available on the World Wide Web.

Thesis (Ph. D.)--University of Chicago. / From the Botanical gazette, v. 49, February, 1910. Contributions from the Hull botanical laboratory 135. Includes bibliographical references. Also available in digital form on the Internet Archive Web site.