Miombo Ecosystem Research Articles

Mapping Woodland Cover in the Miombo Ecosystem: A Comparison of Machine Learning Classifiers

Miombo woodlands in Southern Africa are experiencing accelerated changes due to natural and anthropogenic disturbances. In order to formulate sustainable woodland management strategies in the Miombo ecosystem, timely and up-to-date land cover information is required. Recent advances in remote sensing technology have improved land cover mapping in tropical evergreen ecosystems. However, woodland cover mapping remains a challenge in the Miombo ecosystem. The objective of the study was to evaluate the performance of decision trees (DT), random forests (RF), and support vector machines (SVM) in the context of improving woodland and non-woodland cover mapping in the Miombo ecosystem in Zimbabwe. We used Multidate Landsat 8 spectral and spatial dependence (Moran’s I) variables to map woodland and non-woodland cover. Results show that RF classifier outperformed the SVM and DT classifiers by 4% and 15%, respectively. The RF importance measures show that multidate Landsat 8 spectral and spatial variables had the greatest influence on class-separability in the study area. Therefore, the RF classifier has potential to improve woodland cover mapping in the Miombo ecosystem.

Application of Remote Sensing and Developed Allometric Models for Estimating Wood Carbon Stocks in a North-Western Miombo Woodland Landscape of Tanzania

Quantifying ecosystem carbon stocks is vital for understanding the relationship between changes in land use and cover (LULC) and carbon emissions; however, few studies have documented the impacts of carbon cycling on Miombo ecosystems. Here, we estimate the amounts of wood carbon which is stored and lost as a result of LULC changes in Kagoma Forest Reserve (KFR) for the periods between 1988 and 2010 using GIS data, Landsat imagery, and field observations. The land cover was captured on the basis of Landsat 5 TM and Landsat 7 ETM. The amounts of wood carbon stored and lost were estimated based on four previously developed allometric models. Spatial analysis of the Landsat images shows that in the year 1988, woodlands dominated the area by covering 32.66% whereas in the year 2010 the woodlands covered only 7.34% of the total area. The findings of the current study reveal that KFR had undergone notable changes in terms of LULC for the period of 1988–2010. It was estimated that the woodlands in the KFR lost an average of 4409.79 tCyr-1. In this study, the amount of carbon stocks stored was estimated to be 21457.02 tonnes in tree stem biomass based on the area (1226.12 ha) that was covered by woodlands. We estimated that an average of 17.79 tCh-1was stored in the Miombo woodlands based on the four models. The efforts to ensure sustainable management of the Miombo ecosystem can contribute to the creation of a considerable carbon sink.

Afroboletus luteolus . [Descriptions of Fungi and Bacteria

Abstract A description is provided for Afroboletus luteolus , which if mycorrhizal, is likely to be an important component of miombo ecosystems, sustaining woodland though enhanced uptake of minerals from nutrient-poor soils and by providing protection from desiccation. This species is eaten in Burundi and in the Central Province of Togo, but not in Malawi, where the Chewa and Yao people consider it inedible in their folk-classifications of nature based on the utility of each organism. Some information on its dispersal and transmission, associated organisms and subtrata, habitats, and conservation status is given, along with details of its geographical distribution (Africa (Benin, Burundi, Cameroon, Democratic Republic of the Congo, Guinea, Madagascar, Malawi, Mozambique, Tanzania, Togo, Zambia, Zimbabwe)). No reports of negative economic impacts have been found.

Habitat selection by the Southern (Mashona) Hyliota Hyliota australis in southern Africa: a stand resolution, hierarchical approach

Habitat selection by Hyliota australis in southern Africa is assessed using newly collated habitat and hyliota distribution data. This is done hierarchically, assessing its occurrence in different types of moist-dystrophic savanna, different types of stands of miombo, as well as in different types of landscape mosaics in which miombo occurs. The considerable variation in miombo ecosystem composition, physiognomy and functioning influence its occurrence at all scales. While available data over such a vast area have proved useful—for example, in identifying the importance of ecosystems dominated by Brachystegia spiciformis-there is a dire shortage of data describing the diversity of miombo ecosystems, which hinders a fuller understanding. There is a need for: (1) a hierarchical classification and (2) relatively fine-resolution mapping of the ecosystems of south-central Africa - together ‘an ecosystem inventory’. This is now urgent given the escalating losses of woody cover. All species that rely on undisturbed, ‘old-growth’ canopy miombo are at risk in southern Africa because of high rates of habitat loss on the Mashonaland Plateau, <1% of which occurs in the conservation estate.

Miombo Woodland Utilization and Management, and Impact Perception among Stakeholders in Zambia: A Call for Policy Change in Southern Africa

ABSTRACT Miombo woodland is important for both wood and non-timber forest products (NTFPs). The most commonly harvested products are timber, firewood, charcoal, fencing materials and materials for making farm tools and household items. The non-timber forest products harvested include mushrooms, fruits, medicines etc. A range of user groups such as timber producers, farmers and charcoal producers are involved in the exploitation of different types of woodland resources. Differences in priority of wood products and poor understanding of the Miombo ecosystem often result in each group attributing the negative impacts on the woodland to the other groups. In particular, timber producers accuse the charcoal producers of causing the depletion/decline of the timber species population which are cut as a by-product of charcoal production. However, there is clear evidence suggesting that timber species tend to do better in charcoal production and areas opened up due to slash and burn. This suggests that charcoal production, slash and burn agriculture and timber production may be integrated in forest management. As a result, there is need to formulate policies that provide for the integration of these different land use types into forest management.

Plant communities and tree species associations in a Miombo ecosystem in the Lake Rukwa basin, Southern Tanzania: Implications for conservation

This study assessed the plant communities and species associations in the eastern Miombo woodlands eco-region of the Lake Rukwa basin southern Tanzania. Information was collected from 288 temporary sample plots in 32 sampling sites established randomly in the area and covering as much variation in the landscape as possible. Detailed analysis of the vegetation identified six tree communities of conservation importance, which included Brachystegia boehmii-Pericopsis angolensis woodland, Julbernardia globiflora woodland, Combretum molle-Sclerocarya birrea-Combretum zeyherii-Acacia seyal var fistula woodland, Bridelia cathartica-Diospyros mespiliformis woodland, Brachystegia busseiPterocarpus tinctorius woodland and Brachystegia microphylla-Isoberlinia tomentosa-Hymenocardia acida- Syzygium owariense woodland. Majority of the described plant communities are typical of Miombo ecosystems dominated by trees belonging mainly to the genera Brachystegia and Julbernardia. A mixture of non-legume species but typical of the Miombo ecosystem, however, dominates two out of six described plant communities. Apparently, there is high variability of plant communities in the Miombo ecosystem of southern Tanzania, an indication of the diverse nature of the Miombo ecosystem, also, an implication of high species diversity. This documentation on the patterns of species assemblages is the beginning of unveiling the diversity of assemblage of species in dry forest ecosystems which occupy a wide area in the region. These associations are important targets for conservation and monitoring vegetation changes in this ecosystem.

Impact of small scale tobacco growing on the spatial and temporal distribution of Miombo woodlands in western Tanzania

This paper examines the impact of tobacco farming on Miombo ecosystem in Western Tanzania. Methods employed include literature review, secondary data collection, data sets generation from satellite images (1984, 1995 and 2000) and fieldwork for ground verification. Data sets on tobacco production were computed to generate hectarage cleared for tobacco farming and tree felling for tobacco curing. Trend of tobacco production over time was determined. Findings demonstrate that between 1975 and 1989 production of tobacco was fluctuating. Hence the size of land cleared shows low correlation (R2 = 0.19) with time (years). From 1990 to 1996 tobacco production and the size of cleared land for tobacco farming and curing increased consistently with high correlation coefficiency (R2 = 0.76) with time (years). Land converted from natural vegetation to cultivated land between 1984 and 1995 was 4.7% compared to 11.2% that was converted between 1995 and 2000. Also, about 7.8% of the cultivated land in 1984 had been, by 1995, regenerated into woodland. Between 1995 and 2000 regenerated land is only 2.4% of the total cultivated land. Total cleared land for tobacco growing and curing is projected to double by year 2016. From this study, it is evident that much as deforestation is high, there is vegetation regeneration indicating the ability of Miombo vegetation to recover. Macro-policies have contributed to the land use changes, and consequently on the Miombo ecosystem. Key words: Miombo, tobacco farming, ecosystem degradation.

The ecology and management of the Miombo woodlands for sustainable livelihoods in southern Africa: the case for non-timber forest products

There is new and increasing emphasis on the contribution of non-timber forest products (NTFPs) to improving the livelihoods of rural communities and therefore the need for sustainable management of forest ecosystems of the Miombo woodlands to ensure the continued availability of these NTFPs. This paper examines and reviews some of the work that has been done in the Miombo ecoregion on the ecology, utilisation and management of the Miombo woodlands. The review points to the richness of the Miombo regions' plant species, with an estimated 8 500 species of higher plants, of which over 54% are endemic. The review also highlights the response of the Miombo ecosystem to fire and other anthropogenic processes and the lack of wide-scale adoption or implementation of the major research findings. The review shows that there is a wide range of NTFPs that significantly contribute to the total household income in forest areas and to the national economies through provision of opportunities for small-scale trade especially in the rural areas. However, there is a concern that commercialisation usually results in overexploitation of the resource by the outsiders who become the major players. It is suggested from the review that the lack of capacity by national institutions to manage Miombo woodland resources could be addressed by using participatory approaches to natural resource management. It is also proposed that for sustainable use and management of NTFPs in the Miombo region, research should focus on issues that specifically address NTFPs. In order to facilitate commercialisation and benefit the economically weak, policies that incorporate postharvesting technical assistance and access to information on markets, technology and microcredit should be developed and promoted.

Analysis of land-cover/use change dynamics in Manica Province in Mozambique in a period of transition (1990–2004)

The land-cover/use change dynamics in Manica Province in Mozambique are analysed, based upon interpretation of LANDSAT images, to assess trends and highlight management suggestions considering current policies in place that include the ownership and participation of local communities in natural resources management and the recognition of traditional ownership systems. Applying an object-oriented approach to compare the year 2004 with 1990, one observes that land-cover/use changes are not distributed evenly in the province. A substantial decrease in areas with trees belonging to the ‘ miombo ecosystem’ occurred and an increase of cultivated areas. The nature of agriculture is changing from shifting cultivation to more permanent types. These changes are found in prevalence on gentle slopes and close to the road network. The shift in equilibrium between the cultivated and (semi-) natural vegetated areas to cultivated areas may reduce the availability of forest products and reduce the free ecological services provided by trees. Prolonged cultivation in the same area represents a critical limit for destruction of rootstock and therefore miombo regrowth potential. Community-based management of the natural resources should thus be careful not to disturb the delicate balance of remaining miombo and give emphasis to the regulation of land-use activities, otherwise the negative impacts of land-cover/use change dynamics may fall disproportionately on the poorest members of society.

Woody vegetation structure and composition along a protection gradient in a miombo ecosystem of western Tanzania

Fully protected areas such as National Parks are often assumed to be the best way to conserve plant diversity and maintain intact forest composition and structure. To evaluate this assertion, we sampled trees in areas with four different levels of protection: a National Park, where the protection level is very high, a Game Controlled Area which allows tourist hunting of big game animals, a Forest Reserve which allows selective harvest of trees, and an Open Area where human populations have unrestricted access to forest resources. All four land management areas had healthy size-class distributions with greater numbers of juvenile trees (2–10 cm DBH) than adults. Surprisingly, mean stem density of trees was highest (947 stems ha −1) in the Game Controlled Area but was lowest (635 stems ha −1) in the National Park. The former had the highest basal area value (24 m 2 ha −1) while the human-inhabited Open Area had the lowest (11 m 2 ha −1). Species richness in the Forest Reserve and Game Controlled Area was significantly higher than in the other areas. The total number of plots with unique species not found anywhere else was lowest in the National. Our measures of forest structure and composition therefore show that fully protected National Parks do not necessarily conserve the greatest diversity of tree species or unique species, indicating that a suite of different types of protection strategies may be the key to conservation of trees in these African dry tropical forests.

Land cover assessment with MODIS imagery in southern African Miombo ecosystems

Global land use and land cover products in highly dynamic tropical ecosystems lack the detail needed for natural resource management and monitoring at the national and provincial level. The MODIS sensor provides improved opportunities to combine multispectral and multitemporal data for land use and land cover mapping. In this paper we compare the MODIS Global Land Cover Classification Product with recent land use and land cover maps at the national level over a characteristic location of Miombo woodlands in the province of Zambezia, Mozambique. The performances of three land cover-mapping approaches were assessed: single-date supervised classification, principal component analysis of band-pair difference images, and multitemporal NDVI analysis. Extensive recent field data were used for the definition of the test sites and accuracy assessment. Encouraging results were achieved with the three approaches. The classification results were refined with the help of a digital elevation model. The most consistent results were achieved using principal component analysis of band-pair difference images. This method provided the most accurate classifications for agriculture, wetlands, grasslands, thicket and open forest. The overall classification accuracy reached 90%. The multitemporal NDVI provided a more accurate classification for the dense forest cover class. The selection of the right image dates proved to be critical for all the cases evaluated. The flexibility of these alternatives makes them promising options for rapid and inexpensive land cover mapping in regions of high environmental variability such as tropical developing countries.

Potential and Cost of Carbon Sequestration in the Tanzanian Forest Sector

The forest sector in Tanzania offers ample opportunities to reduce greenhouse gas emissions (GHG) and sequester carbon (C) in terrestrial ecosystems. More than 90% of the country's demand for primary energy is obtained from biomass mostly procured unsustainably from natural forests. This study examines the potential to sequester C through expansion of forest plantations aimed at reducing the dependence on natural forest for wood fuel production, as well as increase the country's output of industrial wood from plantations. These were compared ton conservationoptions in the tropical and miombo ecosystems. Three sequestrationoptions were analyzed, involving the establishment of short rotation and long rotation plantations on about 1.7 × 106 hectares. The short rotation community forestry option has a potential to sequester an equilibrium amount of 197.4 × 106 Mg C by 2024 at a net benefit of 79.5 × 106, while yielding a NPV of 0.46 Mg-1 C. The long rotation options for softwood and hardwood plantations will reach an equilibrium sequestration of 5.6 and 11.8 × 106 Mg C at a negative NPV of 0.60 Mg-1 C and 0.32 Mg-1 C. The three options provide cost competitive opportunities for sequestering about 7.5 × 106 Mg C yr-1 while providing desired forest products and easing the pressure on the natural forests in Tanzania. The endowment costs of the sequestration options were all found to be cheaper than the emission avoidance cost for conservation options which had an average cost of 1.27 Mg-1 C, rising to 7.5 Mg-1 C under some assumptions on vulnerability to encroachment. The estimates shown here may represent the upper bound, because the actual potential will be influenced by market prices for inputs and forest products, land use policy constraints and the structure of global C transactions.

Forestry Mitigation Potential and Costs in Developing Countries Preface

The forest sector in Tanzania offers ample opportunities to reduce greenhouse gas emissions (GHG) and sequestered carbon (C) in terrestrial ecosystems. More than 90% of the country's demand for primary energy is obtained from biomass mostly procured unsustainably from natural forests. This study examines the potential to sequester C through expansion of forest plantations aimed at reducing the dependence on natural forest for wood fuel production, as well as increase the country's output of industrial wood from plantations. These were compared ton conservation options in the tropical and miombo ecosystems. Three sequestration options were analyzed, involving the establishment of short rotation and long rotation plantations on about 1.7 x 106 hectares. The short rotation community forest option has a potential to sequester an equilibrium amount of 197.4 x 106 Mg C by 2024 at a net benefit of $79.5 x 106, while yielding a NPV of $0.46 Mg-1 C. The long rotation options for softwood and hardwood plantations will reach an equilibrium sequestration of 5.6 and 11.8 x 106 Mg C at a negative NPV of $0.60 Mg-1 C and $0.32 Mg-1 C. The three options provide cost competitive opportunities for sequestering about 7.5 x 106 Mg C yr -1more » while providing desired forest products and easing the pressure on the natural forests in Tanzania. The endowment costs of the sequestration options were all found to be cheaper than the emission avoidance cost for conservation options which had an average cost of $1.27 Mg-1 C, rising to $7.5 Mg-1 C under some assumptions on vulnerability to encroachment. The estimates shown here may represent the upper bound, because the actual potential will be influenced by market prices for inputs and forest products, land use policy constraints and the structure of global C transactions.« less