Online public information systems for the management of natural heritage parks pose a number of interesting design challenges, in the sense that the emotional responses of the two primary stakeholder groups - park visitors and park managers - are not necessarily congruent with one another. On the one hand, both managers and visitors need to know what is going on in the park, since conflicting activities by different groups could pose hazards, and interactions between animals and humans need to be kept under control. On the other hand, as Appleton (1976) points out, hazard is attractive to people, and knowing where to find it may result in increased numbers of parks users taking risks that are contrary to the intentions of the parks managers. In this paper, we discuss the iterative design of an online geographic information system (GIS) called UMa. The goal of the system is to minimize negative emotional triggers common tointeractions on the natural landscape and to encourage sustainable decision making by providing potential park visitor groups with frequently updated data on park usage, including wildlife/animal movement patterns. Of particular concern for park managers is park use by school teachers and students, since encounters between school groups and the ecosystem, and school groups and other visitors carry a high-risk potential. For example, periodic permissions are issued to hunters so that they can enter a park and cull overpopulated herds. Knowing where the hunters are is important, since risk is increased if the same area is being occupied, for instance, by a school outing. In this case, providing information directly may be the best course, since teachers will not intentionally seek out risk for their students. Wildlife and people, however, are a different situation; bears are a good example of a case where risks can be introduced by the information system. In an earlier design, the system would show a small icon of a moving bear, placed precisely on the map. The problem is that some parks visitors are attracted to bears, whether in ignorance of the possible danger or because of it. The result was that rather than warning people from the area, the information was an attraction. With the first implemented version of UMa we are working with park managers and primary school educators to explore two hypotheses. Our primary hypothesis is that by designing a useful and appropriate on-line park management system, we can encourage visitors to make sustainable decisions regarding their park visit - decisions that will minimize certain negative impacts on the park and each other, resulting in positive experiences for the visitors. Our secondary hypothesis is that in order to mediate emotional triggers for different park visitor groups we will need to present information at varying levels of granularity. As the system addresses the needs of hunters, for example, we will have to reconsider how abstract and opaque we make the system's information design.