The breeding colony of Northern royal albatross (Diomedea sanfordi) at Taiaroa Head is the only mainland breeding colony of any albatross species. Although it represents only a small proportion of the total population of the species, the ease of access to the colony has led to it becoming an important scientific and tourist resource. Breeding records for the colony date back more than 60 years, and since 1968 daily observation records have been collected. The purpose of this thesis is to present a design for a spatial information system (SIS) for the collection and management of breeding and environmental data, and demonstrate the advantages of such a system. This design is implemented in the form of a prototype SIS for evaluation purposes. What sets this research apart from other studies that have used geographic information system (GIS) software to assist wildlife management is the amount and complexity of data and the time span over which it has been collected. It is proposed that for these data to have maximum utility for wildlife managers at Taiaroa Head, established structured system design and development methodologies are required to ensure data are efficiently stored and important relationships between data are retained.
Three experiments are carried out to evaluate the prototype SIS and demonstrate the range of analyses that are possible. The first experiment inductively explores the spatial relationship between nest site locations and the slope of the terrain at Taiaroa Head—a type of analysis that is commonly used in habitat requirement studies. The second explores the visibility of nest sites with the aim of assessing proposed sites for the installation of video surveillance cameras. The aim of this experiment was to show how the prototype SIS could be used to evaluate proposed management actions in a manner that has no impact on the colony itself. The final experiment explores spatial and temporal relationships that are implicitly present in breeding data by extracting and describing nest dispersal patterns exhibited by the progeny of one Northern royal albatross over five generations. Finally the design and the implementation of the prototype SIS are critically reviewed. It was found that the prototype was subject to two major problems; it did not implement the relational data model according to the system design specification and hard-copy output from analyses were difficult to obtain. These problems arose from the GIS software used to implement the prototype but they are not insurmountable and need not be an issue in an operational implementation. Despite these problems, the prototype SIS did demonstrate that a spatial information system provided significant improvements in the way data are collected, managed, analysed and used in producing useful management information.
