Methodology:

Surveys:

Oblique photographic surveys were carried out using a Nikon D610 camera with a Nikon 28-400mm f3.5-5.6 lens or a Sony A7II camera with Sony 24-105mm f4 and Sigma 100-400mm f5-6.3 lenses, between the hours of 6:00 and 20:00 local time, from 2023-11-21 to 2023-11-30. Photos were taken in RAW and jpeg formats at a minimum shutter speed of 1/2000 of a second and underexposed to retain highlight detail needed to separate species.

Surveys were carried out on either the MV SG Pharos or a rigid inflatable boat (RIB) at a mean distance of 350m from the shore. Photo patterns were conducted in patterns of horizontally across and vertically from top to bottom/bottom to top.

Stitching:

Jpeg images were stitched using Microsoft Image Composite Editor and counted manually using DotDotGoose. Unstitched photographs were occasionally used to check for errors.

Annotation:

Albatrosses were recorded to species level or labelled unidentified when there was insufficient detail to categorise. Birds were counted twice, and for total counts the unidentified birds were assigned to a species based on the ratio of those identified in the colony. Birds were abbreviated like so: black-browed albatross = BBA and grey-headed albatross = GHA.

Data collection:

The following software was used in this dataset:

Microsoft Image Composite Editor software (v 2.0.3; softonic 2024)

DotDotGoose (v. 1.7.9; Ersts 2023)

Data quality:

Diurnal correction factor:

Counts were corrected for diurnal variation in attendance of non-breeding birds and failed breeders as in Poncet et al (2006, 2017), where 2-hour interval counts from selected colonies at Bird Island were used to standardise counts to 12:00h UTC.