Estimating vulnerability under global change: modular modelling of pests

Global change research and decision-making are conducted in an environment where there are many stakeholders, numerous targets and few resources. This calls for global collaboration and sharing of information on a scale that has not been attempted before. It demands a parsimonious approach to resear ch, with the aim of enabling robust decisions to be made with the minimum o f new information. Our approach to these problems has been to develop gener ic analytical tools, which in turn are used to foster collaboration through joint model development, testing and application. This collaboration is ac hieved through a workshop and networking process to establish 'CLIMEX'- or 'DYMEX' -Modelling Networks, and to extend the information to the wider com munity using the world wide web (WWW) (http://www.ento.csiro.au/research/pe stmgmt/IPMModellingNetwork/index.htm). In this paper we outline key feature s of these modelling approaches and illustrate a 'hypothesis-driven' approa ch to climate-matching, using CLIMEX (http://www.ento.csiro.au/research/pes tmgmt/climex/climex.htm) that contrasts with the usual, statistically based , pattern-matching of meteorological data without consideration of possible mechanisms that limit the geographical distribution. We illustrate the nat ure of a generic and modular simulation model built using DYMEX (http://www .ento. csiro.au/research/pestmgmt/dymex/dymexfr.htm), but emphasize the urg ency for the scientific community to collect the data necessary to build re liable population models. We summarize results and conclusions from a globa l change workshop based on the use of both these software tools. They illus trate the advantages of the proposed approach as a means of building collab orative international research communities, which are able to avoid repetit ion by contributing their modules into a library of functions for sharing w ith other users. (C) 2000 Elsevier Science B.V. All rights reserved.