Spatial Interaction of Agricultural Land Uses and their Impacts on Ecosystem Service Provision at the Landscape Scale

Abstract

The relationship between agricultural land use and it impact on ecosystem services, including nutrient cycling and biodiversity conservation, is extremely complex. This complexity has been augmented by isolated research on the impact of agriculture land uses on the landscape's capacity to provide ecosystem services (ES) particularly in most vulnerable areas of Sub-Saharan Africa. Though a considerable number of studies emphasize the nexus between specific land use types and their impact on N-deposition across agriculture landscapes, a sufficient modeling basis for an empirical consideration of spatial interactions between different agricultural land use types at the landscape scale across rural-urbanizing areas in Sub-Saharan Africa is consistently missing. In view of this, the motivation to understand, assess and address significant roles that size, shape, spatial location, and interactivity of different land use patch types play in assessing land use interactions and their impact on ecosystem service provision and the overall landscape resilience necessitated the core of this PhD thesis. This thesis aimed at finding answers to the question of which assessment framework could be employed to understand the interaction of land use types and their impact on ecosystem services, the present thesis introduces a semi-quantitative assessment framework implemented in the GISCAME suite to provide scientific and practical answers to this question. Ahead of the framework development, a thorough review of land use planning documents from selected countries within the WASCAL project area to uncover the key relevance government places on incorporating the ES concept was undertaken. This was with the view that mentioning the concept in such legal document alone does not suggest its relevance if road maps for their implementation is not sufficiently provided, with laid down institutional provisions, roles, responsibilities, support systems and commitments. The outcome of this objective significantly influenced the subsequent objectives of this thesis. <br /> Subsequently, I employed Voronoi tessellation and midpoint displacement algorithms implemented in the Structure Generator (SG4GISCAME) to generate alternative land use mosaics to mimic the patchy agricultural landscape character of the study area. The key objective here was to present the output of this alternative landscape as a partial solution to the data scarcity issue which hinders mapping and hypothetical testing of the landscape structure and their role in landscape resilience. To achieve the objective of identifying core sets of landscape indicators to explore the significant influence of the landscape structure and pattern as an influence on landscape resilience, I employed analytical and statistical multivariate principal component and factor analysis to eliminate the landscape metric redundancy. The outcome helped to propose core set as indicators capable to be used for ecosystem services assessment and land use planning. The result revealed that only 6 landscape metrics had the capacity to explicitly define the configuration and compositional landscape character of the Vea catchment area. This result served a critical input into the development of the assessment framework. In developing a framework to assess the contribution of the landscape spatial structure to the resilience of the socio-ecological system (SES), I mapped the capacity of the landscape structure to provide regulating ecosystem services with the aid of land use maps as proxies. Analytical Hierarchical Processes and Expert stakeholder approaches were used to identify and subsequently map key regulating ES identified from the catchment area. Following, a multi-criteria analysis was employed to link stakeholder mapping and landscape metrics to provide a functional understanding of the interrelationship of both methods and how they provide integrative insights into the landscape resilience and ES trade-off concepts respectively. This assessment was undertaken using a 2012 multi-temporal RapidEye land use classification data and implemented with the aid of the cellular automaton module in GISCAME. <br /> In the absence of explicit ecological modeling and spatial data, the result of this methodology provides a comprehensively rich ES assessment approach not only for the research area, but for transferability across West Africa. The result of this assessment is to inform, across governance levels, different planning, and development scenarios with the potential to alter the landscapes structural character and thereby impede ES flow and resilience of the SES. Indirectly, the relevance of the landscape structure to land use planning was significant across the outcomes of the thesis. Further, the approach establishes potential trade-offs and synergies across the agricultural landscapes structure and thereby suggest planning and management supports to optimize agricultural production and improve ecosystem service flows in Sub-Saharan Africa. Overall, the implementation of the multi-criteria evaluation function in GISCAME demonstrated beyond question, the functional relevance of the GISCAME software tool as the only tried and tested ES integration framework implemented within the WASCAL project area.