Wildlife Population Analysis with GIS: Conservation Management of Royal Albatross

Abstract
This paper examines the use of geographic information systems (GIS) for wildlife management. A prototype GIS that embodies principles of landscape ecology successfully provided useful information that assists with the management of a breeding colony of Royal Albatross in New Zealand. The prototype also provides information that aids in the exploration of the distribution of nest sites within the colony.

There are numerous examples in the literature of the use of GIS for the assessment of habitat requirements and the prediction of suitable areas of habitat for wildlife species. In most cases the use of GIS in this manner have relied on the spatial data operators that are characteristic of GIS for these analyses and predictions. They are typically based on surrogate variables (eg. landcover) that act as predictors of species presence/absence and normally consider a single temporal snapshot.

The adoption of GIS for ecological research and applications has been somewhat slower than that for habitat analyses. Ecological studies focus on organisms and the emphasis has traditionally been on interactions that affect the number of organisms and the changes in those numbers over time. Spatial interactions between organisms, and between organisms and their environment, were either simplified or ignored altogether. A more recent focus on the principles of landscape ecology which considers the spatial and temporal interaction between landscape characteristics and the distribution of organisms would seem to provide a considerable argument for the adoption of GIS. The facilities GIS offer for the storage, management and retrieval of the large quantities of spatially referenced data required for landscape studies make them particularly useful. But there are some obstacles to be overcome. These include:

• While ecological research is rich in theory that relates to processes and ecological functions, it is relatively poor in the spatial concepts and theory that would form the basis of spatial analysis techniques.
• Commercial GIS are weak in the specialised spatial analysis tools required by ecologists; particularly tools for the analysis and description of pattern.
• A lack of spatial population data due to the fact that populations (individuals in a population in particular) are rarely mapped.

This paper describes a prototype spatial information system (SIS) that has been developed by the Department of Information Science at the University of Otago, in collaboration with the Department of Conservation for the collection, maintenance, analysis and reporting of data concerning a breeding colony of Royal Albatross at Taiaroa Head, on the Otago Peninsula. This colony is somewhat unique since it has been the subject of a continuous scientific monitoring and protection effort since the late 1930s. As a result, approximately 60 years of spatial and aspatial breeding data have been collected. The prototype system integrates these data with physical topographic, cultural and natural feature data obtained from aerial and GPS surveys in a manner that allows systematic spatial analyses to be conducted that were not previously feasible.

Three example analyses are discussed that demonstrate the system’s capabilities for the exploration of relationships. This leads to a better understanding of the spatial and temporal relationships between breeding birds and their physical environment and that assist in the management of the colony itself.

Analysis of slope
An experiment is described that examines the extraction of slope information from the spatial database and analyses the interaction between nest locations and this variable. A number of methods for the calculation of slope were considered and rational for the selection of a ‘best’ method are explained. Statistical methods were used to describe the interaction of slope and nest location and this information was used to derive a habitat suitability map for slope. As well a providing a means for carrying out the types of habitat analysis commonly found in applications of GIS for wildlife studies, the prototype system provides capabilities for integrating data synthesised in this manner with an existing attribute database, for subsequent analysis.

Visibility aalysis
This experiment demonstrates the use of spatial information for assisting in the management of the albatross colony. The popularity of the colony as a tourist attraction is placing considerable pressure on the public viewing facilities available. In order to continue to provide a rewarding experience to visitors while minimising the impact on breeding birds, the Department of Conservation decided to investigate the installation of remote video monitoring. The prototype system was used to generate visibility information from an elevation model for a number of proposed camera sites. Results of this analysis will be presented and discussed.

Population dispersion
An experiment will be described that explores the spatial relationship between locations where an albatross chick is raised and locations it chooses as nest sites on its return to the colony as a breeding adult. From one ‘seed’ bird, five generations of its descendants will be analysed. This process involves iteratively trawling an attribute database to locate records attributable to the antecedent bird. Spatial relations are constructed between related records. Results of this analysis are presented for visual interpretation and geometric properties of the spatial relationships are quantified.

The prototype system is significant in that the successful management of the colony is information critical. The albatross population at Taiaroa Head is growing steadily but has yet to reach a level at which it could be considered naturally self-sustaining. The success of the colony therefore depends on the quality of the decisions made regarding its management. The colony has considerable conservation value as it provides a unique opportunity to closely study a species of seabird that is otherwise inaccessible. It’s presence also generates a substantial amount of tourism revenue. The prototype also represents an ecological application of GIS technology that is data rich. Detailed data is maintained for all individuals in the population over a continuous time period at a high spatial scale.