In most regions there are many possible ways of combining numbers of sites into reserve systems that represent a range of natural features. The 'irreplaceability' of a site is operationally defined as the percentage of these alternative systems in which it occurs. This is a fundamental measure of the conservation value of the site in terms of its potential contribution to the achievement of a reservation goal or, alternatively, the options for reservation that are lost if the site is lost. The measure allows decisions to be made on the future of individual sites in the context of their value, in combination with other sites, to the conservation of the full range of natural features in a region. It also provides a logical framework for the design of whole systems of reserves, with decisions proceeding from the most to the least irreplaceable. Irreplaceability can be measured directly for small data sets but must be predicted for regional data sets. A promising approach to prediction is discussed that requires validation with more extensive trials. The irreplaceability of a site depends on a specific reservation target and changes as some of the site's features become progressively represented in reserves elsewhere. The concept of irreplaceability undermines notions of conservation value that are static or based on a single static system of sites to achieve a reservation goal.

Rust (Tranzscheliu discolor) is the most important disease of French prunes in New South Wales (NSW), Australia. Infection by this fungus occurs only when leaves are wet. The length of time for which the leaves are wet and the temperature during such periods are major determinants of whether infection occurs and its likely severity. The use by prune growers of new rust management strategies depends on knowledge of infection events and therefore requires ready access to environmental data from orchards. To satisfy this need, microprocessor‐based device called a Prune Rust Infection Predictor (PRIP) was developed. The instrument measures air temperature, rainfall, and the presence or absence of free water tree canopies. Measurements are taken at 5‐min intervals during wet periods. The instrument calculates whether or not conditions have been suitable for rust infection and potential severity of infection. The system has been operated by advisory staff of NSW Agriculture Fisheries since the 1988/1989 growing season. Potential infection period and related environmental data are collected by personal computers via permanently installed telephone lines from PRIPs located within prune orchards. Prune growers are then advised by recorded telephone messages and given appropriate disease mangement advice. Cumulative potential infection period values are used to provide a prognosis of current‐season rust epidemic severity. An expert system is also being developed to assist growers in making decisions on specific rust management options.

Christi. He has a BS in Computer Science and Mathematics and is currently pursuing a Master's in CS. He has been a USAA intern for three years and has accepted an offer to work as an IT analyst/programmer for USAA upon graduation. His interests are in biometrics and information assurance.