Systems modeling for agroecology and land restoration

• Energy systems language is used to model agroecological land restoration. • Simulations match the observed land-use patterns at farm-scale. • Model describes farmers-mediated ecological succession dynamics. • Local knowledge is key to developing agroecology as an alternative to agribusiness. • The approach deepens the understanding of agroecosystem and land systems dynamics. At the time of the development of H.T. Odum’s energy system theory, the field of agroecology was not yet defined as a scientific discipline. Nevertheless, minimal models were already proposed to describe what today are referred to as agroecological and land restoration practices. In this work, we review literature from the early 1970s to nowadays by tracing a red thread to connect the original formulation of the energy system language with the current understanding of agroecology and land restoration. In the light of this picture, we draw a general application of the energy systems language and modeling to describe land use and land restoration dynamics. We apply this scheme to model and reproduce the land use dynamics of a real restoration project led by a farmers’ family in Mato Grosso do Sul, Brazil. The case study of Sitio Luciana shows the transformation of a monocultural and partially degraded land into a biodiverse, food-producing area by developing a complex agroecosystem owing to human work and farmers’ local ecological knowledge. As an application of the energy system language, we build a dynamical model that closely reproduces observed GIS-retrieved patterns ( global RMSE ∼2% ), highlighting human-mediated ecological succession—targeted at restoring Atlantic Forest—as key to agroecosystems development, and offering scientific validation of the farmers’ local ecological knowledge. This work shows that the energy systems theory and modelling approach, as inherited from H.T. Odum, can: 1) deepen the understanding of local agroecosystem and land system dynamics, including human management; 2) inform the development of non-linear models grounded in both scientific and local knowledge; and 3) offer conceptual guidance for land management and policy strategies.