A generic tool for optimising land-use patterns and landscape structures

We present a flexible and easy to use genetic algorithm-based library for optimising the spatial configurations of land-use. LUPOlib, the (L) under bar and-(U) under bar se (U) under bar attern (O) under bar ptimisation-(lib) under bar rary, can be applied to a variety of spatial planning problems to derive target-driven scenarios that identify trade-offs between conflicting objectives and solve optimum allocation problems (e.g. allocation of reserve sites or management actions). A major novelty is that spatial changes are performed according to a patch topology that allows to simultaneously integrate changes of different landscape elements (e.g. in agricultural fields and linear changes along corridors). The objective function evaluation is based on a grid representation of the landscape where neighbourhood dependencies like lateral flows or the landscape pattern can explicitly be considered. A parameter file allows the user to control the optimisation, the modelled land-use changes, objective weightings and constraints as well as input data. Only the case study-specific objective function needs to be specified in the source code. LUPOlib has been applied so far in two case studies to find optimum trade-offs between habitat requirements of three different bird species and to analyse cost-effectiveness of management actions for species conservation. The results suggest that LUPOlib can be a useful tool to support management decisions. It could be used as an extension to a GIS and for spatially explicit decision support tools. (c) 2007 Elsevier Ltd. All rights reserved.