This communication discusses considerations for the establishment of a spatialized computer model formalizing a prospective of the environmental impacts of public policies regarding land uses. This objective, once applied in West Africa, needs to face several challenges specific to the sub-region: on the one hand, the difficulty of creating a set of data strong enough for enabling the modelling process and on the other hand, the integration of the high variability of its environment, both biophysical (rainfall first) and socially (land tenures first). MOTS-CLÉS : modélisation, modèle conceptuel spatialisé, Afrique de l'Ouest, usages des terres, politiques publiques sur le foncier.
Copyright © by the paper’s authors. Copying permitted for private and academic purposes. Proceedings of the Spatial Analysis and GEOmatics conference, SAGEO 2015.
1.1.
In West-African countries, as in many developing countries, the successful formulation and implementation of land use policies are often hampered by the fact that whatever the issue, we usually do not have enough knowledge about their impacts on global scenarios, especially those related with sustainable development, because of:
The non-availability of good data of the present-time situation, from the biophysical dynamics occurring on the present time to the cadastre itself, but also the weight of uses acting on any piece of land, let aside the temporal accuracy of these data; The absence of scientifically-based but locally adapted tools to process this information: for instance, assessing the use of a piece of land without integrating the multi-ownership and right of use occurring in West Africa on a global scale implies many oversimplification to force the west situation to fit with western standards; In this paper, we propose modelling tools in this context to process this data in order to obtain useful information for planning and making decisions in terms of sustainability, taking account that both data in quality and quantity are considered as enough. 1.2.
Building tools needs a clear understanding of reality. It means scholars tending to build such tools need to search the reference system structure and its functionality. Our departure point is defined by three macro-systems: Nature, Life and Society. Each of them contributes with elements to the structure and from their properties they are defined the inter-relations or it is possible to identify those that are not specified.
The second step looks for describing the state of the system, operation which is
made with help of a subset of variables identified in the system and that constitute
our useful set of variables called "relevant variables". At this level we must know
limitations and restrictions with which we will deal for our purposes. These
limitations and restrictions arise from the simplification of reality, that is, from the
model formulation.
To be able to consider the various facets of a problem at hand, modeling has proven
to be an appropriate tool to provide and combine the necessary information
(
The objective of this paper is to propose a conceptual framework for the multi-scale integrated representation and modeling of land use policy impacts in West-Africa. Specifically, our main aim is to propose a common and shared representation of land use and land management systems in West-Africa for assessing the concurrent impacts of different land use and land management practices on major ecosystem services and disservices, such as carbon and nitrogen storage, water quality and availability, soil degradation, biodiversity losses, and greenhouse gas emission. We consider also social aspects such as poverty reduction and food security, and economic benefits such as income generation.
Practically, many stakeholders are concerned by the land use and land management issues. These stakeholders have not the same understanding and the same point of view on the real objects of the system. Their points of view can be complementary or opposite. However, it is necessary to integrate these various points of view for a better understanding of the target problem. For that, we need a common framework to provide a common representation of the target system but also for knowledge integration from different sources. We need to consider that three fluxes occur in systems: Matter, Energy and Information. On their basis, it is possible to establish relationships among the elements and with the environment. 2.
The research approach is based on multi-dimensional and multi-level approaches.
The objective is to target the land use dynamics at different scales while taking into
account the social, economic, bio physical and policy dimensions. In order to
consider the stakeholders involvement, the conceptual model is based on the general
framework for analyzing sustainability of socio-ecological systems developed by
Copyright © by the paper’s authors. Copying permitted for private and academic purposes. Proceedings of the Spatial Analysis and GEOmatics conference, SAGEO 2015. We seek to provide a common framework for sharing the different representations of land use systems in West-Africa and for knowledge integration from different sources. The general framework for analyzing sustainability of socio-ecological system is based on the assumption that the socio-ecological system (SESs) are composed of multiple subsystems and internal variables within these subsystems at multiple levels. This formalization should concern various land use practices regarding agriculture, forestry, pastoralism, water management, etc. The objective is to assess the competing impacts of different land use practices and policy scenarios on environment (soil degradation, biodiversity loss, carbon & nitrogen storage, water quality and availability, CO2 emission, etc.), society (poverty reduction, food security, health, inequality, etc.) and economic benefits (crop yields, incomes, etc.). That will be achieved by accounting the interactions between social, economic and biophysical dimensions at field, farm, village and region scales (Figure 1).
Using the conceptual model, a meta-model is developed for representing the land
use and land management systems in West-Africa. Several specificities do occur in
West Africa, comparing to other locations in the world:
• Climate, and more specifically rainfall, is not only varying along latitude, i.e.
spatial discrimination. It varies according to seasons such as the majority of
locations in the world. It also vary a lot in time and space between and within
years: this high spatial and temporal intra-annual and inter-annual rainfall
variability
The land tenure problem in West Africa is indeed extremely complex: there are
a lot of different tenure statutes and their definitions are difficult to establish,
particularly for villages in conflict
We based our figures upon the UML representation symbolism (Unifed Modelling Language). The use of the Unified Modeling Language (UML) as the "official" mode of representation of modeling procedures is becoming widespread. Numerous research documents use this language in their communications with other researchers. However, this use has not spread outside of the model makers' community. We however plea for a better formalization and a stronger understanding within this community and between this last and scholars belonging to other sciences and we then stand on this procedure.
In this study, we are concerned by different land use sectors: the agriculture, water, forest and pasture. The objective is to assess the concurrent impacts of different land use sectors. As we consider the land use and land management systems as socioecosystem, we are interested in the interactions between the two ecological and sociological systems at different scales of representation. The ecological system in our conceptual model is represented by the soil, animal, crop, grassland, tree and water components, their own dynamics and interactions. They represent the majors resources used and impacted by the climate and land use and land management in West-Africa.
• •
The animal component concerns only the cattle, sheet and boat that have the most important impacts of land use and land management. The crop component represents a range of species. The tree component represents also a range of species.
The social system represents the human societies, the economic activities, the social organization and the institutional context. Farmers are the main stakeholders concerned by this study. Farmers belong to different socialorganizations on which depends in part of resources and the information access. In addition to farmers, stakeholders concern the public policy agent, local authorities, advisory services, farmers organizations, nongovernmental organizations, research institution, projects , enterprises. In conclusion, the governance system of land use and land management in our conceptual model is composed of farmers, Government Organization, NonGovernment Organization, Farmers Organization, Enterprise, 3.1.2.
3.1.2.1. Multi-dimensional and multi-level interactions The conceptual model takes account four level of description: plot, farm, local and regional scales. Specifically, we seek to analyze the feedbacks loops between these different scales and the governance system that exists at each scale (Table 1). 3.1.2.1. The social interactions In order to represent the interactions between the different stakeholders, three major interacting communication systems are defined. The communication system represents a network that defines the types of relationships between the stakeholders and their interactions. In our study, the communication system is composed by the social, the institutional and the market networks: 1. The social network describes the social relationships and interactions between farmers through different social groups (ethnic, family, neighbourhood). In the social network, the interactions concern exchange for seeds, labour, land, manure, etc. ; 2. The institutional network describes the relations and the interactions between organizations (Government Organisation, NGO, Farmer Organization, Enterprise, Research Institution, etc. ). The institutions interact between them to coordinate their activities and use the institutional network to support the farmers’ activities; 3. The Market network: It will include both price variations and connections, i.e. practical, social and legal restrictions to the free theoretical market (customs, family and social networks, impacts of market sizes and distances).
Based on the general conceptual framework defined, a generic and integrated metamodel is developed to define a common and shared representation of land use and land management in West-Africa (Figure 2). The meta-model is composed of different components of land use and land management system. That concerns the meteorology data, the land and land cover specification, the spatial organization of the targeted system. In addition, the meta-model defines the socio-economic characteristics of the households (Figure 3), the policies supports to households (Figure 4) and their activities (Figure 5). The meta-model integrates the land management practices and their socio-economic impacts. The land management concerns the crop, water, livestock and forest management. Finally, a biophysical component describes the soil and plant characteristics.
For example the household meta-model defines the typology of household from social, economic and policy point of view. From the social point of view, it defines the type of household characteristics: number, farm family size, farm size (ha), livestock size, external income, cropping system, number of external workers, the socio-professional networks. In addition, the meta-model describes the family composition: the groups age, the birth and death rates, labor provided by each group age, money need, female proportion in the group, migration rate of the group, etc. From the economic point of view, the meta-model describes the occupation of the Copyright © by the paper’s authors. Copying permitted for private and academic purposes. Proceedings of the Spatial Analysis and GEOmatics conference, SAGEO 2015. households (farmers, off-farm work) and their activities (crop production, livestock production and forestry activity). From the policy point of view, the meta-model describes the different types on interventions (aid, credit, crop insurance, Off-Farm Wage labor, etc.) that household can beneficiate from the policy makers, project developers, etc.
Copyright © by the paper’s authors. Copying permitted for private and academic purposes. Proceedings of the Spatial Analysis and GEOmatics conference, SAGEO 2015. p S m O e E d a G ca SA d , n e a c e en ta r iv fe r n m O re E p G
Copyright © by the paper’s authors. Copying permitted for private and academic purposes. Proceedings of the Spatial Analysis and GEOmatics conference, SAGEO 2015.
According to
Our issue is intrinsically linked to the characteristics of West African locations:
these belong to complex societies and ecosystems, whose components are linked
together by various interactions, each related to different scientific disciplines. One
cannot select a single interaction among all possible ones as the main and
determining relationship that conditions the evolution of such systems. These
interactions have therefore to be considered as a whole, without which one cannot
understand the functioning of this society. An ABM approach is de facto justified
because such models are adapted to the context, the concepts and the question:
• ABMs are adapted to the context of "developing" countries: The low
availability of quantitative data and the difficulty of launching statistically
reliable investigations limit the use of statistical or multi-criteria
optimization models. ABMs allow incorporating more easily partial results
coming from different sources (agronomic research stations, statistical
questionnaires, open interviews, experts interviews, direct observations or
theoretical behavior rules)
The agent-based approach can integrate quantitative and qualitative
Copyright © by the paper’s authors. Copying permitted for private and academic
purposes. Proceedings of the Spatial Analysis and GEOmatics conference, SAGEO
2015.
variables and thus better formalize the combination of disciplines
They are relevant through their approach, i.e. a differentiation of accesses
to production means and to the related gains. Micro-economic models rest
on the explicit choice of an entity but cannot deal with the effects of the
interactions between several dozen entities because they are rapidly
submerged by the utility functions with several hundred unknowns. Thanks
to their emergence capacity, ABMs can be used according to a very
empirical approach
And finally, we consider as obvious the fact that, for answering a land use issue, the model we consider should be spatialized. Thus, in the ABMs' typology of Gilbert et al, (2006 p.149-152), we stand in the category of spatialized and empirical models.
Simplification vs. Empirism: a practical tension resides between explorers of
processes and explorers of reality. The typology of
Exploring vs. decision support: Along the inherent objective of any model to explore
the variability of a reality through its virtual formalization, modelling for West
African issues implies that it has an applicable objective, i.e. for a development
issue. Practically, such an objective could not be overcome once one should
understand that any funding for research programs do have links with development
institutions, which aims and goals are to mitigate issues related to developing
countries. We support the idea that an ex ante structured formalization of the
hierarchy and the sequence of the tasks to be assessed can mitigate such a constraint.
Modelling interdisciplinary processes: beyond the traditional mantra for
interdisciplinary, the regularity of the failure of interdisciplinary modelling projects
questions the capacity for human scientists to overcome inherent paradigms and
points of view belongings to each discipline, i.e. scientific crispation, and affective
and institutional brakes to such processes, i.e. position crispation. We do support as
well the idea that an ex ante formal brainstorming along resilient methods of
conceptual sharing such as ARDI
Integrated assessment modelling of environmental impacts of land use policy 13