Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Titre : Modelling pastoral mobility to accommodate pastoral land use in land administration : a case study of the Isiolo area, Kenya.

Auteur : Alosa Mulianga, B (Betty)

Etablissement de soutenance : University of Twente International Institute for Geo-Information Science and Earth Observation (ITC)

Grade : Master of Science in Geo-Information Science and Earth Observation (2009)

Résumé
Pastoral land use involves mobility in space and time in search of temporally variable rangeland resources for herds. Such movements are prohibited across privately registered cadastral objects, unless by consent from the land owner, leaving pastoralists marginalized by land rights. This study was undertaken in Isiolo area Kenya, an area characterized by diversified climatic conditions, land uses and land tenures. The overall aim of this study was to accommodate pastoral land use in land administration (LA) through seasonal land rights (SLR). Remote Sensing and GIS tools were used to investigate relationships between temporal pasture fluxes and pastoral seasonal migrations. To achieve the research objectives, multitemporal SPOT NDVI and GIS data were utilized together with field data which was obtained through semi structured questionnaires and participatory mapping. Unsupervised classification of SPOT NDVI data using ISODATA clustering algorithm delineated the land cover into 9 spatial mapping units. Time series analysis on the 9 mapping units realized spatiotemporal NDVI patterns which revealed the bimodal dry and wet climatic seasons that characterized the pastoral migration pattern to and from key dry and wet season grazing areas. GIS animal movement and tracking analyst tools simulated temporal movements of pastoralists based on the migratory routes whose ends defined the seasonal grazing areas using start and end dates of the migration. A GIS multi-criteria decision making method (MCDM) investigated most suitable locations for seasonal land sharing (SLS) among the non pastoral land uses where dry season grazing occurs. Soft systems methodology utilized unified modelling language (UML) to develop conceptual models which integrated the suitable areas for SLS and stakeholder behaviour with the cadastral system. Uniqueness of the developed models was the temporal aspect introduced through SLR which lacked in the existing cadastral system. Other driving forces that influenced the migratory pattern besides demand for herbaceous pastures were deduced from field findings as : access to water, land rights, proximity to grazing sites, security, diseases, pests and cost of grazing. Results of the study presented a strong correlation between the temporal NDVI fluxes in land cover and the spatiotemporal migration of pastoralists. The most suitable areas for SLS included specific locations within the farming, ranching and forestry land uses. The developed conceptual models demonstrated conditions for successful SLS between pastoralists and non pastoralists through SLR in space and time. Comparison of the conceptual models with the real world concluded that SLR would be achieved by public participation, documentation of the migratory corridor, sensitization of stakeholders on benefits of SLS, establishment of regulations to govern SLS and reengineering of land acts. Conclusively, pastoral seasonal migration was mainly influenced by seasonal variability of pastures and SLR was proposed to accommodate pastoral land use in LA. Implementation of SLR however requires inclusion of cadastral information in suitability analysis and instruments to govern the documented migratory corridor.

Mots clés : pastoralists, remote sensing, GIS, spatiotemporal, multi-criteria decision making, soft systems methodology, SPOT NDVI, property rights, seasonal land sharing, seasonal land rights

Version intégrale (ITC)