Multicriteria evaluation (MCE) is a well-tried and effective procedure for structuring and aiding complex decisionmaking processes, especially those involving environmental considerations. Formal deliberative processes have also been successful in aiding understanding and meeting consensus in complex and difficult decision problems which involve more than one decisionmaker. Here, both approaches are combined in a new technique called ‘deliberative multicriteria evaluation’ to assist a group of natural resource managers to decide on a suitable option for recreation and tourism activities in the upper Goulburn–Broken Catchment of Victoria, Australia. This approach is an attempt to combine the advantages of MCE, providing structure and integration in complex decision problems, with the advantages of deliberation and stakeholder interaction provided by a ‘citizens' jury’. An important outcome of the process was the discovery of some crucial aspects of the decision problem that required deeper understanding and assessment if that preferred strategy were to have the desired results. Some suggestions for improving the process are provided but, in general, the stakeholder jury was regarded as a helpful and useful procedure by the decisionmakers and one which aided them in their understanding of the issues of a complex decisionmaking problem.

Summary: Mowing influences two endangered butterfly species, Maculinea nausithous and Maculinea teleius, directly through egg destruction and larval mortality on the mown plants and indirectly through altering the abundance of their sequential resources in meadows (Sanguisorba plants for oviposition and early larval development and Myrmica ant nests for later larval development and pupation). Although conservation biologists have argued that mowing during the adult stage is detrimental to population persistence, it is not obvious how the timing and frequency of mowing impact on population dynamics. A simulation model was used to investigate how current ‘traditional’ mowing regimes could be altered to reconcile butterfly conservation with agriculture. The key mechanism affecting the impact of mowing on population persistence was the interaction between density‐independent and density‐dependent mortalities in different larval stages of each life cycle. Because of this interaction, optimal mowing regimes for butterfly conservation were sensitive to the type of density regulation displayed by each species, and to landscape attributes such as the influence of climate on resource availability and the level of parasitism. Despite this sensitivity, we were able to identify robust mowing regimes appropriate for a wide range of landscape attributes and to derive general management recommendations. Synthesis and applications. Our results showed that the ‘traditional’ mowing regime (twice per year with the second cut during the flight period) was always detrimental to the two butterfly species at both local (single population) and regional (metapopulation) scales. However, mowing once a year, or every second or third year, before or after the flight period, was appropriate for both species in the considered landscapes. Maculinea teleius could persist only at a regional scale, assuming dispersal among the meadows, whereas M. nausithous could persist at both local and regional scales. Thus it is essential that the recommended mowing regimes are applied across several connected meadows within reach of dispersing butterflies if both butterflies are to be conserved in a region.