Paper short abstract:

Using a framework that couples agroecosystem, agent-based small-scale agriculture, and landscape evolution models, we explore the environmental footprint of prehistoric agriculture in the western Mediterranean. Model output is the potential range of human impacts on erosion and land cover.

Paper long abstract:

We present a modeling approach that couples agroecosystem, agent-based small-scale agriculture, and landscape evolution models to explore the environmental footprint of prehistoric agriculture in the western Mediterranean. Dynamically linking these models makes it possible produce concrete and quantifiable estimates of the landscape consequences of different subsistence strategies, agricultural practices, and demographic regimes under various climate scenarios. We use this modeling approach to explore the relationship between past climate, landscape and land cover, and prehistoric agriculture under climate scenarios that characterize the extremes of Mediterranean climate (warm/wet and cold/dry) during the Holocene. We adapt LPJmL (the Lund-Potsdam-Jena-managed-landscape model, (cf. Bondeau et al. 2007)) to the modeling of past agricultural productivity, employ a purpose-built NetLogo agent-based model of early western Mediterranean agriculture, and adapt a published GRASS GIS landscape evolution model. Calibrating these models for past crops and agricultural practices and using a combined downsampling and simulation approach to produce high spatiotemporal resolution paleoclimate data, we simulate realistic potential land use scenarios under past climatic conditions. We here discuss this process with reference to a case study in Provence, reviewing the methodology and data requirements for modeling past agricultural practices and examining the potential range of variability in anthropogenic impacts on erosion and land cover under distinct climate and subsistence scenarios.