You are here: Home Our Work Areas of special importance for the Antipodean Albatross (Diomedea antipodensis) in the Tasman Sea

Areas of special importance for the Antipodean Albatross (Diomedea antipodensis) in the Tasman Sea

submitted by Ben Lascelles, Lincoln Fishpool

Figure 1: Map showing areas of regular use during different life-history stages, and the location of the Tasman Sea area of biological significance which falls in the high seas (beyond national jurisdiction).

Satellite tracking data have been used to identify sites that may qualify as Important Bird Areas (IBA) for the Antipodean Albatross during different life-history stages. As the IBA and EBSA criteria overlap in key ways, we describe an area of importance in the Tasman Sea and the methods used to identify this site. We suggest this approach could also be used to inform the identification of seabird EBSAs.

Download this report

Tasman_Sea_final.pdf — PDF document, 2791Kb

Scientific Background

The selection of Important Bird Areas (IBAs) is achieved through the application of quantitative ornithological criteria (Fishpool & Evans 2001) grounded in up-to-date knowledge of the sizes and trends of bird populations. The criteria ensure that the sites selected as IBAs have true significance for the international conservation of bird populations, and provide a common currency that all IBAs adhere to, thus creating consistency among, and enabling comparability between, sites at sub-regional, regional and global levels. The BirdLife IBA Programme allows for the identification of sites during any part of a species’ life-cycle, and has already identified a variety of sites during the breeding, migration and non-breeding periods. In this respect IBA criteria and EBSA criteria have considerable overlap, therefore the data and methodology used to identify IBAs can be used to inform identification of EBSAs.

Diomedea antipodensis is one of the largest seabirds on earth, and a member of the great albatross (Diomedea spp.) group. It is endemic to New Zealand, breeding on Antipodes Island (4,635-5,737 pairs), the Auckland Islands group (5,800 pairs on Adams, Disappointment and Auckland), Campbell Island (c.10 pairs), and Pitt Island in the Chatham Islands (one pair since 2004). In 1998, it was estimated that there were approximately 39,000 mature individuals; more recent estimates put this at 25,260 mature individuals. Declines in adult survival, productivity and recruitment are largely due to bycatch in longline tuna fisheries. It is currently listed as Vulnerable by IUCN (BirdLife International, 2009a).


Figure 2: All journeys (n = 173) of Antipodean Albatross tracked using Platform Transmitter Terminal (PTT) devices, attached to 42 birds from Antipodes Island (107 tracks in red) and 30 from Adam Island in the Auckland Islands (65 tracks in blue). Image courtesy of the Global Procellariiform Database. Data provided by Kath Walker (Department of Conservation, New Zealand) and David Nicholls (Chisholm Institute, Australia).

Data from satellite tracking (see figure 2) indicate that birds from the Auckland Islands (subspecies D. a. gibsoni) forage mostly west of New Zealand over the Tasman Sea and south of Australia. Those from the Antipodes Islands (D. a. antipodensis) forage east of New Zealand in the South Pacific, as far as the coast of Chile, and have a larger overall range. The tracking data also show that during different life-history stages birds utilize different areas. To afford full protection to this species it is therefore essential to identify areas important for each life-history stage both within territorial waters and on the high seas, i.e. a network of sites.

How the area of special importance to the life history stages of the Antipodean Albatross in the Tasman Sea was identified

Of the global IBA criteria, two (A1 and A4) are currently particularly applicable to seabirds in the marine environment. The IBA criteria explicitly mention the demonstration of “regular use” at a site as a necessary requirement for qualification as an IBA. Whenever possible it is important to utilise data from multiple years to provide justification for this. It has been proposed (BirdLife 2009b) that a regularly used hotspot identified solely on tracking data should be required to meet the following conditions:

“Areas visited by birds from more than one site or during different periods (seasons or years)”


Figure 3: Density distributions for Antipodean Albatross tracked using Platform Transmitter Terminal (PTT) devices, attached to 42 birds from Antipodes (107 tracks) and 30 from Auckland (65 tracks). The Figure shows the 25, 50, 75 and 100% Utilisation Distributions. Image courtesy of the Global Procellariiform Database. Data provided by Kath Walker (Department of Conservation, New Zealand) and David Nicholls (Chisholm Institute, Australia).

Figure 3 shows areas of greatest use for the tracked Antipodean Albatross population over the course of the study. However, in order to identify sites that are important for the different life-history stages of this species, it is necessary to define the nature and timing of the life-history stages, to divide the tracking dataset accordingly, and to define areas of importance for each stage.

Breeding Season Stages

Return to colony







Dec – Feb



May - March the following year

Jan-March (both the following year)

Non-Breeding Season Stages

Successful Breeders

Unsuccessful Breeders


March – Dec (the following year)

Jan/May - Dec

All year (for 5-7 years)

Table 1: Showing dates of different life-history stages of the Antipodean Albatross. Adapted from Brooke (2004)

To determine which areas within the Antipodean Albatross tracking dataset met these conditions of ‘regular use’, the data were split according to the life-history stages outlined in Table 1. This allows identification of sites of importance during both the breeding season (consisting of pre-egg, incubation, brood-guard, and post-guard periods) and non-breeding season. Data were also split for the two different breeding colonies (Antipodes and Auckland islands) because the areas visited by them are different. Within each life-history stage, data were only used if they were available in at least two different years and for a minimum of 4 birds in each year. This is because smaller sample sizes can have an undue influence on the overall areas identified. For each year a Kernel Density Estimation (KDE) was applied, Utilisation Distributions (UDs) were calculated, and the 50% UDs were used to represent core areas of activity. The same process was applied for each life-history stage and each year with sufficient data (see figure 4 for examples). These layers were then combined to create a map of areas of most regular use (see Figure 1). Such areas of special importance in the Tasman Sea (Figure 1) can be seen to be important for multiple life-history stages (seasons) in multiple years.


Figure 4: Areas identified as being of most regular use by Antipodean Albatross tracked from the Auckland Islands, during the incubation period (top left); during the brood-guard period (top right); during the non-breeding season (bottom left) and across all life-history stages and years (bottom right). Image courtesy of the Global Procellariiform Database. Data provided by Kath Walker (Department of Conservation, New Zealand) and David Nicholls (Chisholm Institute, Australia).

This example shows that seabirds often congregate at predictable hotspots (see Weimerskirch 2007 for further example), and linking a hotspot to specific static and/or dynamic environmental variables is an important method for determining a sites’ spatial and temporal stability and/or variability. Linking the sites to environmental variables can clarify whether the importance of an area is determined topographically or bathymetrically (e.g. X km either side of the shelf break, or the presence of a seamount) or by environmental variables (e.g. specific sea surface temperatures, currents, or chlorophyll concentrations). This helps define the features that may be in need of protection to sustain the population. Linking sites to these variables will help to determine for which other taxa and habitats the area might also be important (e.g. Edgar et al. 2008)).

Walker and Elliott (2006) found that birds tracked from both Auckland and Antipodes preferred to forage at the outer edge of shelves and over seamounts, particularly where there were strong currents or eddies and productivity was enhanced, as well as over deep water. The relationship between these variables would be better understood via habitat modeling which may also help to make extrapolations/predictions to non-surveyed areas (e.g. Clarke et al. 2003; Yen et al. 2004).


Sources of data

For this illustration, we used tracking data obtained from the Global Procellariiform Tracking Database held by the Global Seabirds Programme of Birdlife International. Data for Antipodean Albatross were provided to the database by Kath Walker (Department of Conservation, New Zealand) and David Nicholls (Chisholm Institute, Australia).

The Global Procellariiform Database holds tracking data information provided by 57 scientists from 11 countries on 28 species of albatross and petrels. Up to the end of 2008 it held 3,764 tracks obtained from Platform Transmitter Terminal (PTT) & Global Positioning Satellites (GPS) totaling 957,148 hours at sea, as well as 721 tracks obtained from Geolocators (GLS) totaling 61,832 days. A complete analysis of the Procellariiform database would reveal a number of EBSAs on the high seas.

Tracking datasets exist for numerous other seabird species, pinnipeds, cetaceans, turtles and pelagic fish, and the methodologies outlined here could be applicable to these other datasets. (see e.g. Tagging of Pacific Predators (TOPP),, Falabella et al 2009)

To conduct the most complete tracking data analysis requires an adequate sample size of tracked birds within each life-history stage. Most seabird species do not have a complete dataset covering all life-history stages; however for many some form of between-years or between-seasons comparison would be possible.

Important considerations

Identification of biologically significant areas for the various life-history stages of wide-ranging species is inherently linked to the distribution patterns of marine biodiversity and productivity. Seabirds are one of the best taxa for this purpose, because as top predators they are excellent indicators of the state of the wider marine environment, they are easily observed, readily identified and widely surveyed and monitored. In addition they are often linked to a specific range of habitat features, and thus identifying important sites for seabirds is likely to include areas and habitats of importance for a wider range of taxa.

Interpreting tracking data is of vital importance in identifying a potential EBSA on the high seas. Care needs to be taken in analysis of data to ensure that a consistent and comparable approach is used to identifying and delimiting sites. The ecology of individual species needs to be considered at all stages of an analysis to ensure that variation is accounted for in the sites identified, and that the species’ ecology is amenable to a site-based approach during each life-history stage.

Representative sample sizes are especially important for identification of potential EBSAs based solely on tracking data. Bootstrapping (e.g. Manly 2006) may be a useful method for determining if the sample size of tracked birds is representative of the wider population. Adequate sample sizes are likely to vary greatly between species and geographic regions. For studies with small sample sizes, pseudo-replication can be an issue, as the foraging behaviour of a single individual on a single trip can produce hotspots in regions not frequented by any other individual from the same colony or dataset (Seamen et al. 1999). The possibility of missing hotspots should also be borne in mind when interpreting maps, irrespective of the sample size.

Final boundary delineation of EBSAs is always likely to be a difficult choice, and deciding where the boundary should be should include regional and species-specific expertise. It may ultimately be guided by the threats present and the management needs of a given area.

Document Actions