Modelling land-use change for Central America, with special reference to the impact of hurricane Mitch

Land-use systems are highly complex, and any modelling effort of land-use change should account for the complexity of the system. Furthermore, when the aim is to produce realistic scenarios, land-use models should be both spatially and temporally explicit. Using an example of such a model, the conversion of land-use and its effects modelling framework, we explore near-future land-use changes in Central America. Besides a Base and an Optimistic scenario, which assumes yields to approach the present maximum in the region and assumes export and import to increase, focus is on projection of the long-term effects of an extreme event. Scenario assumptions are based on actual data that became available after hurricane Mitch struck the continent. Resulting maps of the Base and Optimistic scenario demonstrate how slow and gradual changes at the national level translate into highly dynamic patterns of land-use change when allocated spatially. Hot-spots of change prove relatively insensitive to changes in income. Particularly land-use change patterns of the most common land-use types, pasture, annual crops, and natural vegetation, differed between both scenarios. The results of the Natural Hazard scenario for Honduras separately as well as for Central America as a whole clearly indicate that the effects of a hurricane on land-use patterns, though initially strong, are likely to largely disappear within a period of 10 years. Concepts from ecology regarding complexity as developed by Holling are used to illustrate the behaviour of the Central American land-use system. Practicabilities for policy makers are indicated, but similar studies and spatially explicit models are needed from sociology and economics to complete our understanding of the long-term effects of an extreme event.