Participant Research Presentations

Each of the participants in the African-GAIM Modelling Workshop brought to the meeting a large body of knowledge and research results. In order to best integrate the ongoing research throughout Africa with the modelling excercises at the workshop, each participant made a 10-minute presentation of his/her own recent research results. The breadth of topics covered highlighted the fact that there is a wealth of research expertise throughout Africa, and that GAIM and IGBP can benefit greatly from enhanced interaction and international collaboration.

DATE (March)
Abdellaoui Abdelkader
Land-use changes in urban environment
Akin Adewale
Monitoring climate and anthropogenic impacts on the hydrological environment
Ayije-Lo Ajavon
Trace gas profiles over tropical Africa
Abigail A. Amissah-Arthur
Land use changes in West Africa: Evolution of crop-livestock production systems in the moist savanna
Rachid Boukchina
Modelling of Nonpoint-Source Pollution From AGRO-Ecosystem: Case of Nitrate
Geoffrey M. S. Chavula
The impact of climate change on water resources of Malawi: Vulnerability assessment and adaptation strategies
James Gambiza
The Regeneration of Zambezi Teak Forests After Logging: The Influence of fire and Herbivory
Willy S. Goma
Quantification of influence of land use on hydrology and atmospheric water balance of the Zambizi river catchment
Chemist Gumbie
Kinfe Hailemariam
Impact of climate change over Awash river basin
Suman Jain
GHG emissions in Zambia from energy sectors in Zambia
Alec Joubert
Equilibrium and transient climate change scenarios over southern Africa
Jenesio I. Kinyamario
Soil carbon and impact of climate change on grassland production
Kwame Mensah Foster
Land use and forest cover monitoring using satellite data
N. Emmanuel Ngwa
The use of remotely sensed data in monitoring local community efforts in managing baseline natural resources in western Cameroon
Michael Ocaido
Developing predictive models for ticks and tick-borne diseases in mixed game and livestock areas around lake Mburo national park in Uganda
Placide Corneille Oke
Climatic change and human health
Daniel O. Olago
Late quaternary glacial-interglacial cycle of climatic and environmental change in Mt. Kenya
Kefa V. O. Rabah
Regional carbon cycle modeling for Africa
Eston Sambo
Developing Strategies for Management of the Lake Chilwa Basin (An Inland Drainage Lake) in Malawi
Mogodisheng B. M. Sekhwela
Woody biomass utilization and harvesting aspects in Botswana: Can these be modelled?
Lerato Thahane
The development of an Integrated Range Condition and Prederred Management Density Model for Thekudumae District, Northwest Province, South Africa
Mame Demba Thiam
Set-up of a GIS for Drought Monitoring in the Northern Part of Senegal (1954-2015)
Otlogetswe Totolo
Assessment of soil conditions in the Gumare area
Pius Yanda
The effect of soil erosion on soil quality in the semi-arid environment

Land Use Changes in Urban Environment

Dr. Abdelkader ABDELLAOUI

Dept. of Geography; Sc. And Tech Uni. Of Algiers; B.P. 32 El Alia 16111 Algiers Algeria

phone: (213) 2 95 91 28; fax: (213) 2 51 58 65


Cities are the sites of the most intense alteration of the natural environment due to land use and other anthropogenic activities. In developing countries, the population increases very quickly; inducing several needs (shelter, clothing, food). This, in turn, causes a permanent variation of the environment. In Algeria, for example, the population was 10.2 million in 1962; it was 16.9 million in 1967, 23 million in 1977 and 28 million in 1987. The growth rates, measured in ten year intervals, rose to 3.8% from 1967-1977, 3.1% from 1977-1987, and finally 2.3% from 1987 to the present year. Although this percentage has decreased over time, the statistic still reflects an elevated rate of growth. Thus, resulting in a permanent apparition of new urban centers along with a rapid extension of existing cities.

To examine the effect of urban expansion on the environment, two examples were studied. The models included the urban complex of Blida which is situated on an agricultural zone in Mitidja, 50 km from the Mediterranean sea, and the oasis of Laghouatt; situated 400 km south of the Mediterranean coast in the semi-arid region. Before 1962, Blida encompassed what is now known as Blida, along with many other small cities. Since that time, the city has expanded 1055 hectares. These hectares were previously tagged for agricultural use. The oasis of Laghouat was originally built around gardens and city; yet due to it's expansion, there is almost no agricultural land remaining. To study the spatial extension of urban space we combined the traditional data (maps) with satellite images; yet several problems arose: 1) in residential districts, the spectral response appeared to be equal to that of vegetation. This can be attributed to pixel heterogeneity. To resolve this problem we characterized the relationship between individual pixel constitution and spectral response. 2) In non-concentrated urban space, confusion in distinguishing between buildings, bare soil, and agricultural space was found. To resolve this problem we used an algorithm based on pattern recognition. 3) In presaharan and mountainous areas there was confusion in distinguishing between buildings and rocks. To resolve this problem we developed an urban index based on the stretching and spectral responses of buildings.


Monitoring climate and anthrophogenic impacts on the hydrological environment

Akin Adewale


March 3-12 1997 Mombasa, Kenya

One of the most important advantages of remote sensing techniques over contemporary traditional techniques in data acquisition is monitoring capability. A number of remote sensing applications have been carried out in hydrological sciences including surface water resource inventory.

In this study, an investigation was made into the application of this remote sensing monitoring capability in assessing changes and trend of change in the measures of the surface water resource of the Sokoto River drainage basin. This occurred during the years of 1956, 1975 and 1986. This basin, representative of drainage basins in the semi-arid lands, is characterized by climatic variation and inadequate data bases. This has not permitted an efficient planning of its water resources.

Using two year anniversary data Landsat MSS imageries (1975, 1986) coupled with the 1956 topographic map as the base, an inventory and monitoring of the upper Catchment of the Sokoto River basin has been attempted at a reconnaissance scale. A decline of about 50% in the drainage parameters of the basin between 1956 and 1975 was obtained, while for 1975 to 1986, the decline was less than 20%. Also a rate of decline of 0.6% per annum in the drainage basin's linear characteristic has been obtained. The various indices of the drainage basin's linear characteristics that have been determined over the three periods (1956, 1975 and 1986) were consequently analyzed using the Analysis of Variance (Anova) statistic. This was done with the aid of a statistical software - `Microstat' on a personal computer.

The impact of the 1967-1985 Sahelian drought vis-à-vis that of the Bakolori reservoir and irrigation projects established in 1978, over the basin were considered in the analysis. It was observed that the Sahelian drought has a significant negative impact on the basin. The impact of the Bakolori reservoir and irrigation projects is difficult to ascertain as to positive or negative with any statistical significance. The results of the Anova analysis for the data showed significant changes between 1956 and 1986 vis-à-vis the stream length over the basin. A computed ratio of 8.336 was obtained as compared with a theoretical `F' ratio of 6.94.

Based on the above analysis (Anova results), the null hypotheses of Nos. 1 and 3 were rejected while the null hypotheses Nos. 2 and 4 were accepted as valid as put forward in the course of investigation.

It is expected that the efficiency of this technique, as well as data from this study will provide useful input into semi-arid water resources modeling schemes.


Ayite-Lo Nohende AJAVON

Atmospheric Chemistry Laboratory

Faculte des Sciences, Universite du Benin, B. P. 1515 Lome, TOGO

The tropospheric concentrations of most reactive chemical species have been found to vary in space and in time all over Tropical Africa. Two categories of chemical species play a major role and are most involved in Africa Tropospheric Chemistry: Trace gases and Aerosols:

The high concentration and the large scale increase of tropospheric ozone observed from data of field measurements campaigns performed in Tropical Africa indicate fundamental change in the chemical behaviour with perturbation of oxidant cycles in Africa troposphere, leading to important photo chemistry and climate change.

Tropical Africa atmospheric phenomena impacts are going to play an important role in global climate for several reasons:

Data from collected rainwater analysis reveal high acid values comparable to those of industrialized countries. Formic and Acetic acids from biomass burning account for more than 50% of tropical African rain acidity.

If data from field measurements campaigns can provide very useful information about trace gases and atmospheric chemistry in Tropical Africa, we must recognize that they remain scarce and very far from representing the amounts of chemical species involved by anthropogenic activities.


Land Use Changes in West Africa:

Evolution of Crop-Livestock Production Systems in the Moist Savanna

Abigail Amissah-Arthur

International Institute of Tropical Agriculture, Ibadan, Nigeria


West Africa contains the largest area of arable farm and grazing systems, suitable for cropping, in the region. On this land, traditionally crops and livestock have been functionally linked, but operationally separated enterprises. However, many of the traditional exchange relationships between pastoralists and crop farmers are disappearing (Powell and William, 1994). Increasing human populations on a finite land base combined with long term climate changes are transforming the specialized systems, based on extensive shifting cultivation and grazing, to more integrated and intensely managed enterprises. Improved crop-livestock production systems are currently developing to satisfy the growing demand for food and prevent the degradation of the natural resource base. According to Winrock International, the greatest opportunity for sustainable increases in agricultural productivity in sub-Saharan Africa lie in agricultural intensification through the evolution and maturation of a mixed crop-livestock farming system.

Development strategies that aim to raise the productivity of specific mixed crop-livestock systems must carefully consider the stage of development of the target area in relation to intensification, and the nature of crop-livestock interactions. Although crop-livestock production systems are evolving as dominant systems, the future imperative of intensification of agriculture indicate a need to develop technologies to meet the demands of these new systems of food production. One perspective on appropriate technology development for crop-livestock production systems is to adopt a resource management approach where land is considered the most critical resource. The land must be managed to optimize crop and livestock returns in a sustainable manner. Improving natural resource management in the moist savanna demands a better understanding of not only the complementary and biological factors (i.e. crop residue for animal feed, and deposition of manure on cropland for soil fertility maintenance), but also the non-complementary interactions involving competition for land, labor, capital and communal resources. Beside these interactions which on their own have important implications for resource management, there is a need for understanding the process by which farmers gain access to, and use natural resources for producing crops and livestock. By examining the manner in which land-use, feed, and soil fertility are managed in the production of crop and livestock the proposed study aims at providing insight into the factors that guide farmers' resource use and management decisions and the manner in which resource use and management are evolving in response to biophysical and socioeconomic change.




Rachid Boukchina

Institute of Arid Lands - Nahal 6051 Gabes Tunisia

Agriculture contributes to non-point source pollution of water resources through leaching and runoff of crop nutrients, pesticides, animal wastes, and through soil erosion from cropland. Development of effective solutions to water quality problems resulting from agricultural land-use requires an appropriate multi-disciplinary approach involving interaction between the components and process of the natural environment and specific human activities associated with land use.

We have developed a numerical multi-disciplinary approach in view of simulation nitrate output from agriculture watersheds through a conceptual model. The simulation of nitrate nonpoint-source pollution requires the coupling of a nitrogen cycle in agriculture ecosystem with a hydrological distributed parameter model. For this, we have used the CEQUEAU (Morin et al, 1981) hydrological model. The CEQUEAU model takes into account the spatial physical variability of the watershed through subdivision in cells of square areas. For each cell, the model simulate the soil water budget. A further cell subdivision into partial cells or fields according to sub-basin is required to allow the determination of the down stream routing from one field to the next. The nitrogen cycle subroutine integrated to CEQUEAU model includes the major process governing nitrogen inputs (organic and mineral fertilizers and atmospheric deposition) internal transformations (mineralization, nitrification and immobilization) and outputs (plant uptake, denitrification, leaching and runoff exportation) for each field over the watershed. Using two year-data from monitoring agriculture watershed (78 hectare), the obtained results showed that the model was able to simulate the seasonal variation of nitrate concentration in runoff accurately.


The regeneration of Zambezi teak forests after logging:

The influence of fire and herbivory.

James Gambiza, University of Zimbabwe, PO Box M P 167, Mt Pleasant, Zimbabwe.

Zambezi teak forests are confined to Kalahari sands and occupy 5.1 percent of the Zimbabwe. These forests contain important timber species such as Guibourtia coleosperma, Pterocarpus angolensis and Baikiaea plurijuga. Concern has been expressed about the poor regeneration of these forests. The impacts of fire and herbivory after logging on the regeneration of the commercial species were being studied. The hypothesis that fire and competition between timber species and other plant species reduces regeneration after logging was being tested.

Four early burning treatments with different fuel load levels were applied to 50 m x 50 m plots just after the end of the 1992/93 growing season. The treatments were: (1) no burning; (2) burning using the natural fuel load; (3) burning after clipping all herbaceous plants to 5 cm (to simulate intense grazing), and (4) burning using twice the natural fuel load. Treatments were replicated three times and applied in a completely randomized block design.

A control (no grazing) and nine treatments involving three times of grazing and three stocking rate levels were applied to one hectare plots in a completely randomized factorial design with three replications. Cattle (preferential grazers) and goats (intermediate feeders preferring browse) were used.

Competition between seedlings of timber species and other plants was studied in an area of 1 hectare. Four treatments were applied to 20 m x 20 m plots in a completely randomized design with four replications. The treatments were: (1) all above-ground shoot material except for the timber species was regularly removed; (2) all above-ground woody material except for the three timber species was regularly removed; (3) all above-ground herbaceous material was regularly removed, and (4) control (no plants removed).

Preliminary indications are that fire and drought are the major factors reducing regeneration of the teak woodlands. Establishment of seedlings of timber species was poor because of drought. The few seedlings (less than 5%) that had germinated were killed by mice. Seedling establishment is apparently an episodic event that is dependent on rainfall.

Time of grazing and stocking rate had significant effects on grass fuel loads. Cattle and goats did not browse timber species. Browse selection patterns were unrelated to chemical composition of the leaves.

A model of the dynamics of logged Zambezi teak forests will be developed. The impacts of large herbivores, fire and frost on key processes such as seed production, dispersal, germination, establishment, mortality, growth rate of different size classes and competition among individuals will be synthesized into a model of the population dynamics of timber species.



Willy S Goma

Department of Meteorology

P.o Box 310095, Lusaka 15301, Zambia

Over the last 50 years, a number of climate anomalies have occurred over the globe in general. Over Zambia extreme events have been experienced resulting in both active and weak as well as good and bad seasons. The much talked about global warming may be responsible for changes in pressure, wind and moisture patterns which can be linked to the observed precipitation anomalies.

To detect changes in rainfall pattern, diagnosis of the rainfall trend over Zambia has been made over selected regions. Mean seasonal rainfall data was obtained from conventional ground based rain- gauges which are spread throughout the country. In addition to rainfall, mean air temperature was also constructed to observe variation in air temperature. The rainfall season ( December, January, February and March) were considered. Correlation coefficients between stations were calculated. Mutually and highly correlated stations identified were grouped together to form coherent regions for analysis.

The observed global temperature records, including ocean and land readings, revealed a warming trend during this century of a magnitude within the range predicted by models (ACTS, 1990). Furthermore, the latter half of 1980 registered the warmest temperatures on record (acts, 1990). In Zambia, as well, mean annual air temperature has risen from around 28 degrees centigrade (1950-1974) to around 29.5 degrees centigrade between (1982-1994) over the region (14.0 s- 18.0s). There was a reduction in mean annual air temperature between (1972-1982) with the lowest being around 26.5 degrees centigrade. The same selected regions show a reduction in mean seasonal rainfall between 1950 and 1978. The mean seasonal rainfall was around 800 mm and after 1978 dropped to 600 mm. The highest rainfalls reported (1950-78) were over 1,000 mm in 1952/53, 1971/2, 1975/76 and 1978/79. While between (1980-1994) the highest was around 800-900 mm during the 1987/88, 1988/89 and 1989/90 seasons. The lowest mean seasonal rainfalls, that fell to around 600 mm between 1950-1978, were in 1949/50, 1961/62, 1963/64, 1971/72 and 1977/78. However, between 1980-1992 the lowest was around 500 mm in 1981/82, 1983/4, 1986/87 and 1991/92.

Moreover, with these few observations, we intend to perform an interdisciplinary meteorological-hydrological research in order to quantify the role of land use-dependent moisture feedback on the atmospheric and hydrological water balance over semi-arid region of southeastern central Africa (Zambia/Zimbabwe). In particular, drought will be investigated to see if land use has contributed to the severity of drought, as they affect the continental moisture recycling. The challenge is to distinguish between two processes affecting rainfall in the Zambezi catchment. First, how far trends in the moisture content of rain producing air masses are explained by decadal and vast land-use changes. Second, the impact of local (subcatchment scale) evapotranspiration on the rainfall distribution on the small scale will be quantified. It is hoped that with attendance of this African-GAIM modelling workshop the project will be a success especially that new software (hydro, Acru, Century) in hydrological modeling have been exposed to the author. (This proposed research will be jointly venture between Institute of Geophysics & Meteorology (German), Delft (Netherlands) and Zambia/Zimbabwe meteorological services.


Impact of Climate Change over Awash River Basin

Kinfe Hailemariam

National Meteorological Services Agency, P.O. Box 1090, Addis Ababa, Ethiopia

Scientific evidence indicates that due to increased concentration of greenhouse gases (GHGs) in the atmosphere, the climate of the earth is changing. A change in climate can alter the spatial and temporal availability of water resources. Therefore, qualitative estimates of hydrologic effects of climate change are essential for understanding and solving the potential water resource problems.

In this experiment an attempt has been made to study sensitivity of water resources to climate change over the Awash River Basin. Awash River starts over the highland of Central Ethiopia , about 150 km west of Addis Ababa, at an altitude of about 3000 m above sea level and ends in Lake Abe at the border with Djibouti Republic, at an altitude of about 250 m above sea level. The total length of the river is about 1200 km and its catchment area is 113700 km2. The climate of Awash varies from humid sub-tropical to arid.

The basin has been divided in to three sub catchments; for a better resolution in calibration & simulation routine. Station based meteorological data were organized into area averages, Grid method GRASS's inverse distance weighing technique is used to obtain area rainfall. Simple arithmetic mean is used to derive area values for the other meteorological parameters. Different sets of temperature & rainfall scenarios were developed using GCM (both transient and CO2 doubling) and incremental scenarios. IASSA integrated Water Balance model (WATBAL) is used to produce runoff scenarios under changed climate.

The model represents water balance streams from the use of continuous functions of relative storage to represent surface outflow, sub-surface outflow, and evapo-transpiration. In the model, the mass balance is written as a differential equation, and storage is lumped in a single mass balance. All components of discharge and infiltration are dependent upon the state variable, relative storage, with exception of base flow which is given as a constant in the mass balance equation. The model contains only three parameters e, a, and Smax which are related to surface runoff, subsurface runoff, and maximum catchment water holding capacity, respectively. The model has been calibrated using 10 year periods (1970-1980), validated for another 6 year period (1981-1986), and then simulated under different climate scenarios.

Results of impact assessment study over the basin showed a decrease of runoff projection by most of the GCM, which ranges from -10% to -34%, with doubling of CO2 and transient period (GFD3, CCCM, GF01), where as UK89 projected a 40% increase. Sensitivity analysis based on incremental scenarios showed a drier & warmer climate change scenario which result in a reduced runoff, with high dependence on the rainfall.




The Zambia Country study on Climate Change under the U.N. convention on Climate Change was carried out by the funding of G.T.Z in two stages: (1) preparation of inventory of GHG emissions in Zambia (2) Abatement costing through possible mitigation options. Using the IPCC methodology, it was estimated that the total CO2 emissions from the consumption of coal and coke in 1990 were 853.37 Gg and 86.5 Gg respectively. The estimated emissions of CO2 from liquid fuels in 1990 were 1859.48 Gt. The estimated CH4 emissions from coal mining in 1990 were 0.316 Gg. Biomass being a potential source of energy in Zambia, GHG emissions from biomass in the form of burning wood as fuel, consumption and production of charcoal were also estimated. The estimates of CH4 and N2O from biomass used as energy in 1990 were as follows

Fuel CH4 (Gg) N2O(Gg)

Wood for direct use 44.68 0.31

Charcoal consumption 0.888 0.05

Charcoal production 29.76 0.04

Although CH4 emissions from biomass are much smaller in weight than CO2 emissions from either solid or liquid fuels, its importance in mitigating the GHG effect should not be underestimated. The global warming potential(GWP) of CH4 is about one or two orders of magnitude larger than that of CO2 and depending on the time frame one chooses. Therefore the above amount of CH4 produces more global warming than the CO2 from fossils. Use of biomass as energy is one potential area which needs to be addressed for any mitigation strategy. This holds true in general for Africa region since biomass is a significant source of energy in Africa.



Alec Joubert

Climatology Research Group,

University of the Witwatersrand, P.O. Wits, 2050, SOUTH AFRICA

Climate change scenarios based on several different general circulation models (GCMs) are provided for Africa, south of the equator (southern Africa). Estimates of climate change are based on both atmospheric GCMs linked to simple mixed-layer oceans and fully-coupled ocean-atmosphere GCMs. The mixed-layer models simulate the equilibrium response of regional climate to an instantaneous doubling of atmospheric carbon dioxide (CO2). The fully-coupled models simulate transient response of the system to gradually-increasing concentrations of anthrophogenic greenhouse gases, and in some cases, sulfate aerosol forcing.

Possible Changes in Mean Temperature

Mixed-layer models simulate an equilibrium response to instantaneously-doubled CO2 of 3-3.5C during the summer of the southern hemisphere (December-February). The transient response to gradually increasing greenhouse gas concentrations simulated by the fully-coupled is 1-2C. Most differences between mixed-layer and fully-coupled models are marked over the central subcontinent south of 15S, and over the southern oceans.

The Influence of Sulfate Aerosols

The direct forcing effects of sulfate aerosols have been incorporated in the Hadley Center for Climate Prediction and Research model by altering surface albedo over several regions of the globe including south central Africa. Over southern Africa, the effect of sulfate aerosol forcing as represented in the model is to reduce both the magnitude and rate of warming predicted for the subcontinent. Based on combined forcing due to increasing greenhouse gases and sulfate aerosol forcing, temperatures may be expected to increase by approximately 0.37C per decade between the present and at least 2060 (a rate which is higher than the global average).

Possible Changes in Mean Rainfall

Both equilibrium and transient climate change simulations indicate that summer rainfall may be expected to increase within the tropics by approximately 10%. Fully-coupled models simulate widespread decreases in summer rainfall over much of the subcontinent south of 15S of 10-15%, whereas mixed-layer models simulate broader regions of rainfall increases (10-15%). However, models of both types display very-little inter-model agreement over southern Africa in terms of the sign of predicted changes. The result is indicative of the problems associated with the representation of rainfall produced by connective processes within GCMs, and it's representation at the coarse spatial scales typical of all GCMs. Currently therefore, little confidence can be placed in estimates of rainfall changes based on GCMs alone.

Possible Changes in Diurnal Temperature Range

There is a marked seasonality in the simulated changes in diurnal temperature range over southern Africa. During the summer and Autumn (September-May), diurnal temperature range is expected to increase, indicating possible decreases in cloud cover, relative humidity, soil moisture, runoff and precipitation (see above). However, during winter (June-August) the diurnal temperature range simulated by the Hadley Center model is expected to decrease, due to a more rapid increase in minimum temperatures, relative to maximum temperatures. This in turn implies the opposite sequence of changes in surface variables to that expected during the summer, and is indicative of the importance of the presence of moisture over southern Africa in the determination of a wide range of climate changes.



Foster Kwame Mensah

The increasing rate of deforestation in Ghana is significantly affected by agricultural land use change in the forest zone. The driving force behind this has been found to be the pressure exerted on land as a result of the increase in population growth over the years.

The existence of geographically referenced data on the spatial distribution of forested areas as well as associated attributes, such as total acreage of different cover types, and forest inventory data is key to effective monitoring of the forest zone. The use of satellite imagery can provide cost effective means for creating spatial data sets for monitoring and for providing inputs for forest zone decision support systems.

The objective of this project is to produce digital and hard copy maps of the forest cover and of land use in a specified buffer zone for two forest reserves in the eastern part of Ghana, at a scale of 1:30,000. This data set is required for use as part of baseline inventory of forest cover which monitors deforestation areas. The spatial nature of this data set clearly depicts the distribution of land use and of forest cover in the forest ecosystem. By realizing that the spatial fragmentation of forest cover has an effect on ecological processes, it is intended to produce forest cover and land use change for the area of interest and to explore methods for linking the cartographic output to ecological models.

The primary data source for this project was SPOT panchromatic and multi-spectral satellite data.



NGWA Nebasina Emmanuel. P.O. Box 12551, Yaounde I University. Rep. of Cameroon.


Rural community use of data derived from remote sensing, such as aerial photographs and satellite images is yielding results and gradually bridging the gap between financial budget managers and administrative officials on the one hand and rural populations in Cameroon on the other. This is because a fair percentage of such officials often does not appreciate the importance of authorizing financial expenditures for research, field work, and related materials. Population pressure on what is considered as free and unlimited natural resources in the study area, coupled with other factors is modifying the ecological systems and rendering them more and more fragile.

This small ongoing project was started by the Rural Geography Laboratory of Yaounde University. The objective is to show that field use of available remotely sensed information and accompanying instruments can quickly augment awareness among local communities and generate data for a follow up of a number of natural resource management activities.

Initiated in 1989, a group of faculty and students went into the field with aerial photos. The group used the photos to demonstrate the negative changes which could be identified in the local areas of state corporation managers, local council mayors, rural co-operative managers, village chiefs, and farm group leaders. These negative effects touched mostly on water resources, woodlots, grazing lands, and so on. Today, the evolution in the various forms of resources is yielding multiplier effects and results. Not only is there motivation to preserve village water catchments with adapted trees, but also beekeeping in these woodlots has been identified by the communities as an additional socio-economic incentive. Farm group leaders have been involved in the ban of phytosanitory products which affect flower when applied with sprayers. This in turn affects bee activities and consequently, the quantity and quality of honey. Cattle breeders are currently collaborating with rural councils in the preservation of wetlands, which hitherto, constituted intensive grazing zones for them in the dry season. Their own reactions are coming in by way of identifying and classifying grasses for a traditional pharmacopoeia from which they are traditionally treating cow skin diseases, cow diarrhea, pulmonary tract infections, etc. These socio-economic multiplier effects from each action are bringing in specialized NGOs and technical ministries which continue to upgrade wildlife habitat preservation and enhancement, the maintenance of biodiversity, village woodlots, and medicinal plant preservation. All these can be identified from remotely sensed data. It is hoped that further monitoring would in due course widen the scope and strategies of each rural community in redressing the inadequacies so far identified, and lead to the next step where a model would be put in place for quantification, resulting in a more efficient scenario for managing peasant efforts in the sustainable use of their natural resources. Community leaders would have also moved from their narrow view of issues to a more open approach wherein allocations are made for the acquisition of visual aids such as remotely sensed material for rural mass sensitization.




Lake Mburo National Park is located in the Mbarara district of western Uganda. This park is surrounded by cattle ranches and communal grazing areas. In the areas surrounding the park, wild animals, mainly impala (Aepeyceros melampus ), Zebra(Equus burchelli), eland (Taurotragus oryx ) warthog ( Phacochoerus aethiopicus), and buffalo (Syncerus caffer), among others, are found sharing the rangeland with livestock (cattle and goats). In these areas, the stockholders consider wild animals as a source of ticks and tick-borne diseases. It is against this background that this study was initiated with the major objective of developing predictive models for tick-borne diseases. These models once developed, will be used for designing cheaper optimal methods for controlling ticks and tick-borne diseases so that a harmonious mixed game and livestock production system can be developed in this geographical location.

The study is intended to run for five years. It involves monitoring the dynamics of tick populations in the rangeland and the prevalence of tick-borne diseases with particular emphasis in East Coast Fever (ECF) in cattle in the ranches and communal grazing areas.

The tick population (Rhipicephalus appendiculatus) will be monitored in pasture in relation to variability of rainfall, temperature, pasture height and cover, wild animal population dynamics, and dipping practice (concentration & frequency). This study is being done on two ranches, on one communal grazing area, and on three sites in the park . The tick population models will be designed based on the above variables. Factor analysis and statistical technique will be used to identify those independent variables that can be used for predicting tick populations. Poison regression will be used to analyze the reduced data to either exponential or seasonal local tick population models.

The prevalence of tick-borne diseases is being observed on two ranches and on one communal grazing site. This will be done in relation to tick population dynamics, dipping practice, and cattle disease susceptibility, which is dependent on herd age and breed dynamics. Regression analysis will be used to analyze the data into multiple regression models.



(Weather conditions and cerebrospinal meningitis in Benin )

M. BOKO, P. C. OKE Laboratory of Climatology, UNB / FLASH / DGAT P. O. BOX 03 - 1122 Cotonou Benin

J. P. BESANCENOT CLIMAT & SANTE , Faculty of Medicine, 7 bld Jeanne d'Arc, 21033 Dijon Cedex, France

Harmattan apparently has a great hold over the spatio-temporal distribution of cerebrospinal meningitis (CSM). According to many authors, heavy responsibilities for this lie in the damage caused to surface defense mechanisms by adverse climatic conditions and seasonal respiratory infections.

CSM is a major concern in developing countries, particularly in tropical climates, since overcrowding, poor sanitation, poor living conditions and in some instances, limited access to adequate health care can contribute to large epidemic outbreaks. The disease mainly affects the northernmost regions of Benin, from November to March, April or even May. On an average, acme occurs in February-March. Regression analysis confirmed that a part of the space/time variability of the disease was due to harmattan and dry weather. Mediation of adverse climatic conditions in the infectious phenomenon is therefore indisputable, through both outbreaks of concurrent respiratory tract infections and sudden decreasing immunity. However, the climate-meningitis relationship proved to be weaker than it was usually assumed, perhaps because this relationship is partly eclipsed by both anthropic actions (vaccination campaigns) and latency period before disease outbreak. If attention was for a long time focused on the role of the harmattan, this assertion can be more complex than the classical model several authors have drawn from the persistent heat and humidity of the previous months. The effects of the harmattan are far from wholly recuperative.

The harmattan on its own is not enough to explain the temporal variability of the CSM to be profitably used as predictor. Finally, current knowledge of the climatic factors that determine whether or not meningococcal infection will lead to overt disease so far is too scanty for definite conclusions. The rather obscure influence of time of year and weather conditions on CSM reemphasizes the multi-factorial and complex nature of the etiology of this disease.


Late Quaternary Glacial-Interglacial Cycle of Climate and Environmental Change on Mount Kenya

Daniel O. Olago

Department of Geology, College of Biological and Physical Sciences,

University of Nairobi, P.O. Box 30197, Nairobi, KENYA

The late Quaternary palaeoclimate and palaeoenvironment of Mount Kenya (which sits astride the equator at 37E) are reconstructed from sedimentological and geochemical analyses of a closely-dated, carbonate-free, organic rich lake sediment core retrieved from Sacred Lake on the northeastern flank of the mountain. More particularly, core chronology was established by radiocarbon and U/Th dating techniques, and then stable carbon isotope record was presented. The global palaeoclimatological linkages are established with reference to other late Quaternary palaeo-records, e.g. from other tropical lakes and swamps, polar ice cores and deep sea sediments.

The lake sediment core spans the whole of the last glacial - interglacial cycle. The record shows that the mountain vegetation was dominated by C4 plants during the glacial period and by C3 plants during the Holocene - aquatic vegetation underwent similar changes as the terrestrial component. Trends in 13C discrimination by aquatic microorganisms, particularly algae, paralleled the vegetation trends, with heavier isotope values during the glacial period and lighter values during the Holocene. It is demonstrated that changes in atmospheric CO2 concentrations over the glacial - interglacial cycle probably impacted more strongly on tropical ecosystems than temperature, which has been thought to have the dominant effect. In addition, during the glacial period, the stress effect of low CO2 was accentuated by increased aridity.

Spectral analysis of the stable carbon isotope record reveals the presence of orbital (Milankovitch) periodicities, indicating that changes in the earth's orbital parameters have strongly influenced the tropical palaeoclimate and palaeoenvironment. Furthermore, earth-intrinsic palaeoclimatological influences on Mount Kenya arise from changes in the North Atlantic circulation, and they appear to be particularly associated with the Heinrich events and Bond cycles.



K. V. O. Rabah

Department of Physics, University of Nairobi

PO Box 30197, Nairobi, Kenya

We focus on Africa's present and future contribution to excess atmospheric carbon dioxide by an already existing global carbon cycle model. The logic behind this is to develop a model that represents Africa's responsibility in terms of the net flux of anthrophogenic CO2 into the atmosphere and hence, its contribution to enhanced radiative forcing (the greenhouse effect) and possible consequences (e.g., global warming). See Fig. 1 for the pictorial model.

The following procedure is adopted to help develop the model:

1. Partitioning of the CO2 fluxes by source (Africa, or not Africa) analogous to "flavors" of a gas in chemical tracer model.

2. Replication of the terrestrial biosphere for multiple vegetation types using a global database.

3. Breakdown of the model by continent using ratio analysis to explicitly focus on Africa relative to the rest of the world.

Initially, the ocean will be treated as a global aggregate, interacting with the terrestrial components through the atmosphere as one component. This is done simply to reduce the complexity of handling many variables in the first implementation of the model, but will be incorporated later. We present a model that has been developed to indicate Africa's share of the anthrophogenic perturbation to the global carbon cycle. Our result will provide information to policy makers in Africa and to others involved in the international greenhouse debate.

View Figure 1: Africa Biosphere



E.Y. Sambo, University of Malawi, Biology Department, Chancellor College,

P.O. Box 280, Zomba, Malawi


Lake Chilwa (located in the southeastern part of Malawi) is a shallow, enclosed lake that receives inflow of water from six major rivers. Two of these rivers (Likangala and Mulunguzi) pass through the town of Zomba at the foot of Zomba Plateau. The Mulunguzi river originates from the Zomba plateau and is dammed to provide water for the Zomba Municipality, a town with a growing population of around 70,000. The catchment of Lake Chilwa includes the Zomba Plateau and Mulanje massif. The lake is characterized by large seasonal fluctuations in water levels and has periodically dried out completely, notably in the dry seasons of 1967 and 1995. In good years it has provided a significant proportion of Malawi's fish output. However, major changes in land use over the past 25 years have occurred around the lake and the whole basin, and they include the introduction of rice irrigation schemes on the plains, development of commercial farms (some growing tobacco) further upland and increased human settlements throughout the basin. Lake Chilwa is now a proposed RAMSAR site. Baseline studies were completed in 1996 for accession to the RAMSAR Convention.

Environmental problems

Land use changes (in particular conversion of woodland to agriculture) has caused decrease in land cover. Increasing human settlements in most parts of the basin are causing extensive deforestation which has recently extended to the riverine areas, as a result of extraction of wood. Soil erosion is a big concern in the whole catchment, causing siltation of Lake Chilwa and the mouths of major rivers. Floods have become frequent and more widespread in the Domasi valley partly due to the problem of sedimentation.

Nutrient enrichment of the lake from leached fertilizers from the farmland is expected to cause eutrophication. Pesticides from farmland and domestic/industrial effluent from Zomba town are sources of pollution for the lake. Cyclic complete drying out of the lake is not well understood. This affects fish population and interferes with communication of island people by boat. Population health needs on the wetland (especially regarding control and treatment of bilharzia and malaria) have been one of the main issues.

Management implications

In the implementation of reforestation strategies, attention should be given to involvement of the community. The machinery is already there through Village Forestry Action Committees. Agroforestry initiatives and other tree planting activities are key factors, including special efforts to rehabilitate river banks. However, not much is known about how the communities have in the past managed their natural resources sustainably. These indigenous knowledge systems should be investigated, evaluated and taken into account.

Grazing by domestic animals on the Chilwa plains is an important activity which needs proper assessment as a land use system in the emerging scenario of competing needs. The use of fire on the plains, within the marshes and swamps of the lake (dominated by Typha domingensis) and on the whole catchment occurs almost annually. The Typha swamps are home to over 150 species of both migrant and resident birds, as well as a dry season habitat for roosting of adult locusts, especially when the fires have destroyed the grasslands.

The dynamics of fisheries of Lake Chilwa regulate the level of economic activities of the surrounding community. The drying out of the lake has previously demonstrated a major disruption of the normal way of life but other adaptive strategies have been used by the population. Future considerations should focus on protecting breeding stock of fish and their breeding grounds such as reservoirs at river mouths. There should be proper understanding of the nutrient yields of N and P and the flow of carbon in the system.

Mapping out of flood mitigation strategies will be important. This should be linked to a proper understanding of the hydrological system of the whole catchment, including runoff, sediment yield etc. Monitoring of the hydrological and ecological factors important in the ecosystem will be key to future management e.g. regular monitoring of water quality.


There are a number of questions that can be asked, solutions of which could be explored by modeling. Below, I set some of these out:

The ACRU, CENTURY and STELLA MODELS are well suited to provide answers and future areas of importance for developing management strategies of the Lake Chilwa basin.

Future plans

The Biology Department is developing a management plan for the Lake Chilwa basin. A multidisciplinary team consisting of ecologists, hydrologists, sociologists and others will work with the community towards this goal.



Mogodisheng B.M. Sekhwela

Woody biomass harvesting and use mainly for fuel wood and construction poles are part of the many benefits derived from natural woodlands which normally surrounds settlements. Rural communities in many parts of developing countries have derived products of various kinds and have obtained other non-tangible benefits, such as shade and aesthetic beauty from natural woodlands for millennia. In Botswana where more than 80% of household energy, in both rural and urban areas and where construction poles are sources mainly from natural woodlands, wood cutting is regarded as a major contributor to deforestation. Different activities contribute to wood cutting, resulting in various products which are used in different socio-economic activities found in the rural areas.

The different sources of wood products and end uses can be identified, resulting in the production of flow charts highlighting the movement of wood products. If such relationships can be defined quantitatively, as well as the influential socio-economic activities depended on human population, it should be possible to make a mathematical representation of such interactive relationships. Such models could be used to further explore and increase the understanding of the use and management of woody biomass sources which are continually declining, yet important CO2 sinks. These would assist in the mitigation strategies of likely impacts of global change, particularly sequestration of CO2 which is solely dependent on the existence of photo synthetically active plants.



L. Thahane

University of North West

P/Bag X2046, Mmabatho 2735. South Africa

Range condition and rainfall are some of the most important determinants of available biomass for livestock production in rangelands. Much work has been done in this field in South Africa, but there are big gaps especially in those areas farthest from Universities. This is an example of a range condition model developed for a communal rangeland, in the absence of sites that can be regarded as pristine or benchmarks.

Forty sites were selected in a relatively homogenous grazing area. These sites were selected so as to cover various levels of grazing pressure. Quantitative species composition data was subjected to ordination techniques in the Integrated System for Plant Dynamics (ISPD). The data was then refined using the Detrended Correspondence Analysis ordination. The refined data was then coordinated with centered DECORANA. The first axis corresponded to degradation because of the sampling bias, and the second and higher ordination axes were summed up in the 2nd axis as residuals. The resultant model is considered to be an acceptable fit because the maximum Y-axis value is less than half of the maximum X-axis value.

Schmidtia pappophoroides emerges as a strong indicator of degradation showing an inverse linear relationship with increased grazing pressure. The relationship between S. papppophoroides and degradation gives an r2 value of 0.99. Aristida congesta and Elephantina elephantorrhiza were also significant as degradation indicators even though they showed much weaker correlation. In this area Schmidtia pappophoroides can be used as a single indicator of degradation.

There is a need to develop models for all other unmodeled site types and climatic sub regions, to correlate the different range conditions with biomass production and rainfall variations, and to finally develop an expert system that can be used for monitoring change and to inform range management.



Mame Demba THIAM

This research - development was elaborated since 1994 and it has five years duration. The main goal of this project is to survey :

For this program our expectation is to pay a particular attention to the place of drought cycles recorded since the end of the sixties and to one which should probably concern the Soudano-Sahelian area between 2005-2015.

The diagnosis approach deals with a remote method; so with multiple sources of data, we try a multitemporal analysis exploiting a Geographic Information System (GIS).

In our purpose, the necessary challenge is to propose in an open and advisable mind a provisional definition of desertification. Desertification means the process, which is hardly reversible on a human scale, of mutation of the Sahelo-Soudanian and Sahelian area into some kind of desert area featured by lowest resources and durability.

The location area does concern the biologic Sahelian zone in the Northwestern of Senegal. This is an average of 1/3 of total of the country, which should be affected by constraints like those existing in arid zones, if arrived on a cycle of dryness comparably in the seventies. This reason explains the manageable vision of our problematic project.

We have chosen a test-zone on "Lac de Guiers borders". The analysis is based on a retrospective approach of landscapes with a systematic cartography. This leads to a kinematics atlas. The result helps us to build our model of the test-zone.

We hope that we will produce a prospective analysis by cartographic simulation of the whole area. In pursuing our goal, we are looking for a way to develop a cartographic model of risks during the period 2005-2015 and consequently a territorial theory for the strategic safety of human settlement in concern.

Assessment of Soil Conditions within Eroded Areas in the Gumare Area
(West of the Okavango Delta in Northwestern Botswana)

Otlogetswe Totolo

Environmental Science, University of Botswana

Private Bag 0022, Gaborone, BOTSWANA

The Okavango Delta is an internationally renowned ecosystem and has provided multiple uses to the local people. In the past, the western part of this extremely fragile ecosystem has been put under a lot of pressure by cattle keepers. But of late, there is a lot of interest in assessing environmental conditions, both natural and socio-economic, after the scourge of cattle lung disease which has befallen the Ngamiland District. The aspect of the physical environment which is of interest to this investigation is to assess the soil conditions as they exist now and how they can be improved in order to reduce wind erosion. This is very important more especially that this part of the country has been under enormous pressure from cattle grazing before the massive killing of cattle infected by the cattle lung disease. Most of the soils around Gumare are very silty, and therefore when they are dry, are highly susceptible to wind erosion. This problem is exacerbated by over-utilization of the resources in the area which leave the soils bare like overgrazing. Therefore, it is absolutely essential that the scale of the problem be quantified and understood in order to give well informed advice to the local community and the government. Otherwise we might be faced with a situation where the Okavango Delta might be seriously affected and eventually disappear. In the past the delta was well protected by the tsetse fly and some of it was designated as wildlife management areas. The tsetse fly has been eradicated and therefore the delta is more accessible than in the past.

The aim of this investigation is to:

1. Quantify wind blown deposits in some selected sites using soil traps

2. Assess the aggregate stability of both the degraded area and soils under vegetation cover.

3. To investigate the magnitude of both sheet and wind erosion and assess which is the most problematic.

4. Characterize the soils of the selected sites in terms of their morphological and chemical properties.

5. Work out the possibility of stabilizing the soils in the area in order to reduce the spread of soil degradation into the delta.


1. Identify the badly degraded areas within the study using Satellite imagery.

2. Field work to characterize the soils in the main delineated areas. (This will include soil micro morphology, soil nutrients and clay minerals).

3. Determine the aggregate stability of all the soil samples.

4. Photographs of the different sampling sites will be taken in order to use them as reference points for monitoring purposes.

5. Set up soil traps in order to assess the quality of the soil that is lost.

Linkages to the African Modeling Workshop

After attending the workshop in Mombasa, there are many opportunities with respect to using STELLA, ACRU, Hydro and Century in order to test different scenarios within the study area. Soil loss in the Gumare is of great concern and therefore the eco-hydrologic models could not have come at a better time. The Century model will be mostly used for modeling vegetation production and nutrient dynamics within the system. Opportunities for using STELLA are also immense with respect to coming up with a model which can simulate environmental conditions within the study area.



Pius Yanda , Institute of Resource Assessment,

University of Dar es Salaam, Tanzania

Semi-arid Central Tanzania is among the most highly degraded areas in the country. In 1973 the government of Tanzania launched a land reclamation program. Both physical and biological measures were employed. However, more attention was given to reclamation of gullies. In 1979 livestock was evacuated from the area to allow natural vegetation regeneration. No consideration was made on the spatial variations in the soil degradation and thus requiring different land reclamation measures. Objectives of this study were: (a) to classify different degradation classes, (b) to identify potentials and limitations for crop growing in each of these classes. Methods used were the following: (a) Mapping of land form and soil units, (b) Mapping of spatial variation in soil degradation, (c) Aerial photo interpretation, (d) Soil sampling and analysis, (e) Infiltration measurements.

Landscape in the area was broadly divided into two categories: the eroded and the depositional areas. In the erosional areas, soil degradation intensity classes are based on the properties of the eroded soil horizons. Five degradational units were established and four units were established in the depositional zone.

Results from this study illustrate that the removal of A horizon greatly alters the infiltration capacity of soil surfaces because of the unfavorable soil structure in the underlying B horizon caused by physical and chemical differences between these horizons. The average clay content in the B horizon is 42.5% compared to 25.6% in the A horizon. Similarly, the average organic carbon content is 1.2% in the A horizon as compared to 0.3% in the B horizon. Also the influence of organic matter content in maintaining soil nutrients is limited to the A horizon. This means that erosion of the A horizon leads to a dramatic decline in soil nutrients.

Modelling could be an important tool to assess the severity of soil degradation in the different degradation units (i.e. soil erosion rates, decline in soil infiltration capacity, ultimate effects on crop yields). Also, through modeling, it is possible to assess the influence of each of the soil degradation factors in the study areas. Modelling could as well be used to identify practical and effective soil conservation techniques.

Go back to African Workshop Report