Data and Methodology

This study integrates data from different sources, and uses different methods and approaches. To analyze wildlife and livestock population changes, animal counts were used from aerial census conducted by the Department of Resource Survey and Remote Sensing (DRSRS) between 1975 and 2007 using systematic reconnaissance flights fitted with GPS sets. To characterize land use/cover changes, satellite data was used as outlined in Table 1 below.

Table 1: Key characteristics of satellite data

Satellite Sensor Acquisition date Resolution

Landsat MSS Landsat 2 1975/02/11 75/120 Landsat TM Landsat 4 1986/10/17 30/120 ALOS AVNIR-2 2007/07/24 10

Our methodology is summarized in Figure 2. Based on the results of the unsupervised classification, ground data and authors knowledge of the study area, clusters of pixels representing various land use/cover categories were selected as training sets and their spectral response patterns were generated. Based on spectral separability, the training areas were suitably modified and the final spectral response patterns were generated, and used to classify the images using a Gaussian maximum likelihood per pixel classifier. The classification accuracies were then estimated following the procedure by Bishop et al (1975). Post classification analyses were performed to quantify and identify the changes that occurred over the study period. Further, using GIS techniques, qualitative interrelationships between various variables were analyzed and a conceptual model depicting land use dynamics developed and key relationships evaluated.

4

Results and Discussions

Classified results of land use/cover for 1975, 1986 and 2007 are presented in Figure 3 and the changes in Table 2.

Charles N. Mundia

Table 2: The approach adopted for the analysis of land use/cover changes.

Figure 3 Land use/cover maps for Masai Mara ecosystem for 1975, 1986 and 2007 from multispectral satellite images.

Farmland, grassland, shrubland and forest area were the dominant land use/cover classes. Significant spatial expansion in farmland and the rapid decrease in forest cover within and close to the Masai Mara National Reserve were observed in the 1986 and 2007 land use/cover maps.

Table 2: Observed land use/cover changes for Masai Mara.

Land use/cover 1975 2007 % change Area (ha) % Area (ha) % 1975-2007 Farmland 15,540 1.2 147,490 11.5 10.3 Grassland 243,940 18.9 191,070 14.9 - 4.0 Forest 138,430 10.8 123,600 9.6 -1.2 Shrubland 886,090 69.0 822,030 64.0 - 5.0 Total 1,284,000 100.0 1,284,000 100.0

Analysis of long-term aerial census data show very rapid decline for most wildlife species. Population trends for the livestock and the three most abundant wildlife species in Masai Mara ecosystem are presented in Figure 4.

Sustainability of Tourism in Kenya: Analysis of land use/cover changes and animal population dynamics in Masai Mara

Environmental Issues and Sustainable Development

179 Figure 4: Population trends of wildlife and livestock in Masai Mara National Reserve and surrounding areas.

Source: Department of Resource Survey and Remote Sensing (DRSRS).

Analysis of wildlife dynamics indicates that the wildebeest population decreased by 74% while the expansion of mechanized farming took place on the wet season rangelands, which were fenced off to exclude wildlife. Livestock population trends over the same period shows fluctuating patterns with an increase in recent years. 4.1 Conceptual model

Following remote sensing based analysis and extensive fieldwork assessment in Masai Mara, a conceptual model was developed and shown in Figure 5 that analyzes the dynamics of ecosystem change in terms of competition for land and competition for biomass. The total land area of the ecosystem is in demand for subsistence, tourism, mechanized cultivation, and for grazing both livestock and wildlife. These land demands are controlled by socio-economic factors but also compete for limited space.

The transitions while driven by many factors have underlying human drivers. Foremost among these is land conversion to agriculture especially the expansion of mechanized farming which is controlled by economic factors. Wildlife and livestock compete for biomass and the size of livestock is linked to pastoralist’s decision and their wealth. Around the conservation areas, a significant portion of pastoralist wealth derives from wildlife related tourism activities. All these factors together with the tourism management style and the existing environmental conservation policies have contributed to the current scenario of rapid land use/cover changes and wildlife decline.

Charles N. Mundia

Figure 5: Conceptual model depicting the land use/cove dynamics in Masai Mara Ecosystem 4.2 Conclusions

Increasing land use/cover changes and the drastic decline in wildlife population have been observed in the Masai Mara ecosystem. Information about these changes is important for planning for sustainable utilization of resources. To study the long-term land use/cover changes and wildlife population trends, data was integrated from different sources and different methods were used and approaches including satellite remote sensing, field surveys and wildlife population trend analysis. Post classification analysis and integration of various data using GIS approach was adopted to examine changes and wildlife population dynamics.

The results show rapid land use/cover conversions and a drastic decline for a wide range of wildlife species. It is urgent that all necessary measures are taken in order to strike a balance between wildlife conservation and economic development. It is also important to revise and strengthen the wildlife policy so that wildlife species in private group ranches outside the Masai Mara National Reserve are protected and habitats are conserved to ensure sustainable tourism and balanced economic development.

Acknowledgement

This work was funded by Japanese Society for the Promotion of Sciences References

Jolly, G.M. Sampling methods for aerial censuses of wildlife populations. East African Agricultural and Forestry Journal (Special issue): Nairobi, Kenya, 1969.

Kamusoko, C.; Aniya, M. Land use/cover change and landscape fragmentation analysis in the Bindura district, Zimbabwe. Land Degradation & Development, 2007, 18, 221-233.

Ottichilo, W.K.; De Leeuw, J.; Skidmore, A. K.; Prins, H. H. T.; Said, M.Y. Population of large non-migratory wild herbivores and livestock in the Masai Mara Ecosystem, Kenya. African Journal of Ecology, 2001, 38,

202-216.

Ottichilo, W.K.; Khaemba, W.M. Validation of observer and aircraft calibration for aerial surveys of animals. Africa Journal of Ecology, 2001, 39, 45-50.

Serneels, S.; Lambin, E.F. Proximate causes of land use change in Narok District, Kenya: A spatial statistical model. Agriculture, Ecosystem and Environment, 2001, 85, 65-81.

Touson, C. Challenges of sustainable tourism development in the developing world. Tourism Management,

2001, 22, 289-303.

Ottichilo, W.K., De Leeuw, J., Skidmore, A. K., Prins, H. H. T. & Said, M.Y. (2001) Population of large non- migratory wild herbivores and livestock in the Masai Mara Ecosystem, Kenya. African journal of Ecology. 38: 202-216.

Estimation of Non-Point Source Pollutants in Upper Mahaweli Catchment of

Sri Lanka by using GIS

E. A. S. K. Ratnapriyaa_*

a _{Postgraduate Institute of Agriculture, University of Peradeniya, Old Galaha Road, Peradeniya, Sri Lanka.}

[email protected]

KEYWORDS:

Catchment base, GIS, Non-Point Source (NPS), pollutants, Upper Mahaweli Catchment

ABSTRACT

Upper Mahaweli Catchment (UMC) is the upper portion of the Mahaweli watershed area, above the Rantembe dam and it covers an area of about 3118 km2. Four hydro power plants located inside this catchment that contributes 60% of the hydro electricity supply in Sri Lanka. Therefore, this area is very vital to national economy. Qualities of the surface water of this catchment have been affected extensively due to increase of urban and sub-urban agglomeration along water bodies, wastes disposal by local authorities, soil erosion, and intensive agricultural activities. Identification and estimating point source have been properly done over the years due to simplicity of the process. Eventhough, assessment of Non-Point Source (NPS) is indispensable part of catchment base wastewater management; inadequate studies were done due to NPS pollution is highly correlated with specially oriented data and analysis. Geographical Information System (GIS) is popular for its capability to integrate and analysis layers of spatially oriented large amount of information. This study mainly focuses on assessment of average annual NPS pollutant load contributing from the sub-catchment area of UMC area. Surface runoff and the soil erosion process are two main transport agents of NPS pollutants. In this study, measurements were taken for pollutants in selected points inside the catchment and same points were used to demarcate micro catchments in ArcGIS Hydro Extension. Therefore, those micro-catchments are the contributed area for pollutant measurement points. Soil erosion was calculated by using Universal Soil Loss Equation (USLE) in GIS raster format and hydrological equations were used to calculate the surface runoff per year. Rate of erosion and pollutant concentration in the eroded soil used to estimate the pollutant transport with soil and annual surface runoff and pollutant concentration in water samples were used to estimate pollutant transport by surface runoff. Summation of above the estimated values considered as the amount of pollutant produced by NPS pollutes in the particular micro-catchment per year. Eventhough, the estimated loads might be not the absolute values, it can provide sufficient information to decision makers for implementing proper catchment base wastewater management system.