Quantifying Similarity of Correlations between Seabird and Cetacean Distributions and Environment in the Northeast Atlantic

Electronic versions

  • Rose Greensmith

Abstract

Understanding the role of mechanistic processes in species distributions is a key aspect of understanding the species spatial ecology, particularly interspecific interactions between species with overlapping resource requirements. However, comprehensive understanding is often hindered by spatial and temporal coverage of abundance data and lack of established statistical methodology to derive this from abundance data. This study aims to address these challenges by quantifying similarity among distributions of seabird and cetacean species. Intra-guild or taxa separation could indicate potential habitat partitioning, and equally, similarity between sympatric species could indicate potential coexistence. This study used zero-inflated generalised linear models to model a large-collation of seabird and cetacean abundance data across the northeast Atlantic, so that relationships within their likely ranges can be identified. Clustering and principal component analysis of the conditional model regression coefficients were used to quantitatively identify similarity between seabird and cetacean distributions and their environment within each species likely range. There was dissimilarity within guilds, and similarity between some sympatric species from different guilds. Furthermore, the scale of the relationship between abundance and their environment was distinct between taxa, as non-delphinid cetaceans had much stronger correlations than delphinids and seabirds. Explainers of dissimilarity can be simplified into species’ spatial, behavioural and prey differences. These outcomes align with coexistence and competition theories, indicate that products of mechanistic processes are observable on a large scale, and that interspecific interactions are potentially involved. Future research includes identifying if interspecific interactions are the responsible mechanisms driving this similarity structure, then how to appropriately integrate this in species distribution modelling processes to improve ecological realism.

Details

Original languageEnglish
Awarding Institution
Supervisors/Advisors
Award date26 Oct 2021