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

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 through 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 the 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 non-significant correlation between remotely sense 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.生物多样性调查正愈发广泛地开展。然而，针对此类数据的标准化分析技术尚未成熟。本研究依托生物多样性数据，探讨了多元排序技术（multivariate ordination techniques）在评估物种-生境关系中的应用。研究于美国蒙大拿州的冰川国家公园（Glacier National Park）展开，选取鸟类与蝴蝶作为研究的分类类群。研究团队在覆盖多样生境的区域布设了生物多样性调查样地，并于1987年至1989年间开展持续监测。通过对全部采样样地的物种检出/未检出数据，对物种的常见度与稀有度进行分级。在三年的采样监测周期内，共观测到历史记录中约86%的蝴蝶物种与70%的鸟类物种。本次研究的三年监测期内，各调查样地的物种丰富度呈现显著的延续性。同时，物种多样性较高的样地与稀有物种分布样地之间存在显著的重叠区域。主成分分析（Principal Components Analysis）与聚类分析（cluster analysis）能够较好地辨识物种-生境关系。海拔、样地结构多样性与湿度是解释物种分布格局的核心环境因子。卡方分析（chi-square analysis）同样为物种-生境关系研究提供了重要见解，结果显示鸟类相较于蝴蝶对生境的特异性更强。生境多样性分析结果表明，样地的遥感（remote sensing）光谱分类多样性与鸟类、蝴蝶的物种丰富度均呈正相关，但相关性未达到显著水平。坡向、坡度与海拔多样性与物种丰富度则呈负相关或关联程度可忽略不计。