|Thesis abstract: |
The sustainability of food production has emerged as a crucial issue in the global development context. World population growth and dietary trends have been causing a significant increase in food demand in spite of nearly constant natural resources supply. The uncertainty of future climatic and socio-economic scenarios further challenges our capacity to achieve a sustainable balance between food demand and supply in the long run. This is the major challenge the agricultural sector and policies are facing.
This being the context, the role of farmers becomes pivotal in the sustainable management of natural resources involved in the agricultural production. Their behavior is led by economic and social (e.g., food requirement) objectives, while the ecological sphere is seldom considered in strategic decisions. According to the Millennium Ecosystem Assessment, this lack is contributing to the occurring degradation of the main ecosystems services provided by the natural sphere to the agricultural one: beside food provision, natural ecosystems guarantee several services like water regulation and purification, pest control, carbon storage and biodiversity. The poor knowledge of the mutual interactions between human and natural spheres and the consequent scarce awareness among the actors involved (both farmers and decision makers) further slow and limit the inclusion of ecological considerations in the design of effective management policies. For this reason, tools able to tackle the complexity of coupled socio-ecological systems (SES) are required in order to support the creation of effective policies.
Among the existing modeling techniques, Agent-based models (ABMs) have emerged in environmental sciences as a way to better capture complex system characteristics of coupled SES. In particular, they are able to represent the behavior of human actors more realistically, accounting for bounded rationality, heterogeneity, interactions, evolutionary learning and out-of equilibrium dynamics, and to combine this representation with a dynamic heterogeneous representation of the spatial environment. On the other hand, four major challenges of ABMs in SES have been identified in the literature: (1) design and parameterizing of agent decision models, (2) verification, validation and sensitivity analysis, (3) integration of socio-demographic, ecological, and biophysical models, and (4) spatial representation.
The aim of the current research is to investigate the use of ABMs in coupled SES as a tool for supporting decision processes aimed at creating sustainable resource management policies. In particular, the project will focus on the ecological sub-models, with the aim of capturing the relations between human agents¿ behavior and the main components of the ecological sphere, i.e., ecosystem services. Moreover, proper indicators for assessing the state of the main ecosystem services will be created and integrated among the output of the model, using both field and remote sensing data. The methodology that will be developed will essentially support the analysis of the effectiveness and sustainability of specific agricultural techniques and policy strategies at the system level (e.g., region and/or nation). Summarizing, beyond a deep analysis of the methodological challenges of ABMs in SES, the project will investigate the role of these models as a tool for facilitating the difficult integration of ecological dimension in the creation of sustainable resource management policies of food production systems.
The thesis work plan is organized in the following steps: (I) first year: firstly, the state-of-the-art review of ABMs for SES is carried out; in particular, suitable models of ecosystem services and of the mutual interactions between the human and ecological spheres are investigated (through the analysis of a case study located in the Peruvian Amazon); secondly, proper food security and environmental impacts indicators are investigated (application to a case study in the Gaza Strip); (II) second year: existing algorithms and approaches are tested and, basing on the test results, the proper agent-based sub-model of human component is developed; (III) third year: the ecological sub-model is developed and integrated with the human one, and the whole model is simulated under different socio-economic scenarios, in order to assess selected management strategies (e.g., Payment of Ecosystem Services) under a sustainability perspective.