Multi-scale assessment of distribution and density of procellariiform seabirds within the Northern Antarctic Peninsula marine ecosystem

Summary

This study used ship-based surveys and specialized statistical models to assess the distribution and population density of 11 tube-nosed seabird species in the Northern Antarctic Peninsula marine ecosystem during summer months (2003-2011). The research revealed two distinct distribution patterns: open ocean species concentrated in the southern Drake Passage and coastal species around the Peninsula and South Shetland Islands. Ocean surface temperature and water depth were the most important environmental factors determining where seabirds were found. The models performed moderately to well for predicting presence or absence of species (accuracy scores 0.60-0.80) but were less effective at predicting detailed density variations. The study provides baseline distribution data essential for understanding potential impacts of climate change and human activities, particularly krill fishing, on Southern Ocean seabirds. The research highlights concerns about krill depletion affecting krill-dependent predators and establishes important habitat relationships for future conservation and management decisions.

This map shows the combined predicted density of all 11 procellariiform seabird species, with each species normalized and summed to identify overall hotspots. Warmer colors indicate higher combined density, cooler colors show lower density.

Key Findings

Eleven tube-nosed seabird species were analyzed from 32,181 kilometers of ship surveys conducted between 2003-2011.

Two distinct distribution patterns were identified: oceanic species concentrated in the southern Drake Passage versus coastal species around the Peninsula and South Shetland Islands.

Sea surface temperature and water depth were the most important environmental factors determining seabird distribution across all species.

Models performed moderately to well for predicting species presence/absence (accuracy 0.60-0.80), but were less effective at predicting density variations.

The study establishes baseline distribution data essential for monitoring ecosystem changes and assessing potential impacts from climate change and commercial krill fishing on Southern Ocean seabirds.

Abstract

The Antarctic Peninsula is one of the most rapidly warming regions on earth, and it is likely that the abundance and distribution of marine predators will change as a result. Procellariiform seabirds are highly mobile predators, which target specific habitat characteristics associated with underlying distributions of prey and areas of increased prey availability. We use ship surveys and hurdle models, to estimate the summer distribution and relative density of 11 seabird species within the northern Antarctic Peninsula marine ecosystem. Models differed among species; however, sea surface temperature and depth were frequently associated with seabird occurrence and had the greatest explanatory power across many species. Null models based on observation data were better at predicting seabird density than models that included environmental covariates. This suggests that the main driver of distribution patterns is the broad-scale habitat features, and fine-scale aggregations within these ranges are harder to predict. Our seabird distribution models reflect known habitat associations, species hotspots, and community organization relative to oceanic and coastal marine processes. Application of species distribution models will benefit the assessments of critical habitat and potential responses to climate change and anthropogenic disturbance, which will provide insight into how species may change in polar ecosystems.

Published in

ICES Journal of Marine Science

2021

Authors

Warwick-Evans, V., Santora, J. A., Waggitt, J. J., Trathan, P. N.

Institutions

Ecosystems Department, British Antarctic SurveyFisheries Ecology Division, Southwest Fisheries Science Center, NOAADepartment of Applied Mathematics, University of California Santa CruzSchool of Ocean Sciences, Bangor University

Methods

DataField

Read the full paper

Go to doi.org