Using Biodiversity Data to Assess Species--Habitat Relationships in Glacier National Park, Montana

Abstract

Biodiversity surveys are becoming increasingly popular. However, standard analysis techniques for these data have not yet been developed. This paper explores the use of multivariate ordination techniques for assessing species—habitat relationships using biodiversity data. The research was conducted in Glacier National Park, Montana, and birds and butterflies were chosen as the taxonomic groups of study. Biodiversity assessment sites were established throughout a range of habitats and monitored from 1987 through 1989. Presence/absence sampling over the total number of sampling sites was used to classify species commonness and rarity. Approximately 86% of the historically recorded butterflies and 70% of the historically recorded bird species have been observed in the 3 yr of sampling. During the 3 yr of this study there was a striking continuity of species richness per site. There was also a striking overlap between sites that support high species diversity and sites that support rare species. Principal components analysis and cluster analysis worked well in discerning species—habitat relationships. Elevation, structural diversity of the site, and moisture were the major factors explaining species distributions. A chi—square analysis also provided some insights into species—habitat relationships, showing birds were more habitat specific than butterflies. Habitat diversity analyses demonstrated a positive but nonsignificant correlation between remotely sensed spectral—class diversity of a site and species richness for both birds and butterflies. Aspect, slope, and elevation diversity had a negative or negligible relationship with species richness.