Bioactivity of beaver castoreum constituents using principal components analysis

Abstract

North American beaver (Castor canadensis) were observed to sniff from the water and make land visits to some synthetic chemical components of castoreum placed on experimental scent mounds (ESM). In previous analysis, the elicitation (presence/absence), completeness, and/or strength (number, duration) of these key responses served as separate measures of biological activity. In this paper, we used principal components analysis (PCA) to combine linearly six related measures of observed response and one index of overnight visitation calculated over all trials. The first principal component accounted for a majority of the variation and allowed ranking of the samples based on their composite bioactivity. A second PCA, based only on response trials (excluding trials with no responses), showed that responses to the synthetic samples, once elicited, did not vary greatly in completeness or strength. None of the samples evoked responses as complete or strong as the castoreum control. Castoreum also elicited more multiple land visits (repeated visits to the ESM by the same individual or by more than one family member) than the synthetic samples, indicating that an understanding of the castoreum chemosignal requires consideration of responses by the family unit, and not just the land visit by the initial responder.