Feature

Watching the Albatross

Concern over the loss of albatrosses caught on commercial fishing lines has revealed that we know relatively little about the largest of seabirds.

Susan Waugh and Paul Sagar

New Zealand scientists are studying the populations of albatrosses in New Zealand and looking at their feeding behaviour to gain an understanding of how numbers are being affected by albatrosses being caught up as bycatch in fisheries.

In the six years since the first news of the extent of albatross bycatch in fisheries in our national waters caught public attention, the focus has been on documenting exactly how many of New Zealand's endemic albatross species are being caught on tuna long-line hooks. Of the 24 taxa of the albatrosses in the world, 13 breed at New Zealand sites and ten of these are endemic -- they breed nowhere else in the world. Of the 13 different kinds of albatross breeding here, at least 10 have been identified from specimens returned as bycatch from New Zealand fisheries.

Although bycatch records reveal that there is potential for serious declines in some New Zealand albatross populations, the information needed from the bird populations to say exactly how population size and age-structure are affected has generally been lacking. We know that during the breeding season, seabirds such as albatrosses may feed far from their island breeding sites and can travel thousands of kilometers to feeding areas. However, when and where they go, what they feed on, and to what extent they interact with fisheries are questions which are largely unstudied.

Until recently, the easiest thing to determine was how long they were away from the nest. However, the recent development of satellite telemetry allows determination of the locations of albatross foraging areas and distances traveled.

Without taking a two-fold approach -- investigating both population change and feeding ecology -- making the link between high bycatch of a particular species and decreasing populations is impossible. Additionally, important influences on a population and its food-sources which may result from environmental change, such as the influence of El Nio (Southern Oscillation temperature anomalies and changing sea-surface temperatures) could be underestimated.

Not all fishing activities are detrimental to seabirds. For example, black-browed mollymawks breeding in the Falkland Islands have been shown to obtain about 15% of the food to fed to chicks as offal discharged from a squid fishery. However, it has not been shown whether access to this additional food supply has led to an increase in breeding success or survival to breeding age. Fisheries may also compete for the same food resources as albatrosses by removing the birds' prey, such as squid.

Two studies have been implemented at the National Institute of Water and Atmospheric Research (NIWA) to:

• investigate the population dynamics of three New Zealand sub-Antarctic albatross populations
• examine links with fisheries and environmental factors
• investigate if the birds' foraging behaviour may lead them to interact with fishing vessels.

Both studies started in July 1995 and will continue for at least three years. These programmes are funded by the Foundation for Research, Science and Technology, the Department of Conservation using a levy made on the fishing industry, and the New Zealand-France Cultural Agreement, and are run in collaboration with researchers at DoC, the Museum of New Zealand, and the French Centre National de la Recherche Scientifique (CNRS).

The former two agencies have had a long involvement with seabird research in New Zealand. CNRS pioneered satellite tracking studies, and have completed over 30 years of research on albatrosses breeding on French sub-Antarctic Islands in the Indian Ocean.

Buller's Mollymawk

The first of these studies is of Buller's mollymawk (formerly called Southern Buller's mollymawk), which nests on the bush-covered cliffs of The Snares, 90 kilometers south of Stewart Island, and the Solander Islands, off the southern coast of Southland. Some 10,960 pairs of this endemic albatross breed at these two sites, with an estimated 8,460 breeding at The Snares. The breeding population at The Snares increased about 78% between 1969 and 1992. However, at the Solander Islands, the situation is more complex, with no comprehensive counts being completed until February 1996.

Buller's mollymawks were the albatross most commonly caught by the Japanese southern bluefin tuna longline fishery in New Zealand waters during 1989-1992, comprising 25% of the total number of albatrosses caught. Most were caught off Fiordland, on the Puysegur Bank, and over the Solander Trough west of Stewart Island, during the months April to July. Autopsies of some of these birds indicated that they were adults, with approximately equal numbers of males and females.

April to October is the chick-rearing period for Buller's mollymawk. The co-incidence of bycatch and breeding season raises the question of which breeding population/s are the bycatch birds likely to come from? The Snares or Solander Islands?

Satellite tracking of the birds breeding at The Snares showed that during March (the incubation stage of the breeding cycle) off-duty birds foraged either in the Tasman Sea, as far west as Tasmania, or along the east coast of the South Island, as far north as Kaikoura. From hatching in mid-March until about the end of April, all breeding birds fed at the edge of the Snares Shelf, about 300 kilometers east of The Snares.

Foraging patterns changed again from May to mid-June, when the waters along the east coast of the South Island were favored by all birds. Finally, during July, the latter part of the chick-rearing stage, most birds fed closer to The Snares, staying in the waters immediately south of the South Island, and only traveling as far north as Otago Peninsula; although during this period one bird switched coasts and made several feeding trips to the waters off Hokitika.

By integrating computer databases, we are now assessing the co-incidence of birds and fishing vessels, and examining whether there are any relationships between where the birds feed, sea-surface temperatures, and wind-speed and direction at the time the birds leave The Snares.

Campbell Island Study

The second study is taking place at Campbell Island, the strong-hold of the Campbell mollymawk (formerly the New Zealand black-browed mollymawk) and the only New Zealand breeding site of the grey-headed mollymawk. Grey-headed mollymawks are found breeding at eight island groups around the Southern Ocean, but the Campbell mollymawk is endemic to Campbell Island. The island is 700 kilometers south of the mainland of New Zealand, and supports a population of some 26,000 Campbell and around 6,000 grey-headed mollymawks (7% of the world population).

Detailed work by DoC has revealed that the breeding population of these two species has declined markedly since the first records of their numbers were made in 1942. Between 35-57% fewer mollymawks were breeding on Campbell Island in 1987 than in 1942, when war-time coastwatchers made observations and took photographs of the albatross colonies. The exact timing of the decline and the current trend of breeding numbers are still being studied by DoC. The fall in breeding numbers at the mixed mollymawk colonies was mainly because grey-headed mollymawk were diminishing at a faster rate.

Both of these mollymawks are found in the Japanese tuna long-line bycatch around New Zealand, making up 20% and 7% of the total albatross bycatch respectively between 1989 and 1992. Only Buller's mollymawk is caught more frequently than Campbell mollymawk, and the Campbell mollymawk is also frequently caught in the Japanese long-line fishery in the Australian region. It is a well known ship-follower, unlike the grey-headed mollymawk, which probably explains some of the difference between the bycatch rates of the two species. There is a strong likelihood that a combination of fisheries and environmental factors have contributed to the population changes of these two species.

The albatross studies being carried out by NIWA, and co-researchers are examining how different parts of the study populations have behaved over recent decades. The field teams are banding birds and recording the band numbers of birds returning to the colony. This work tells the researchers how long individual mollymawks live for, and also how many birds on average are surviving from one year to the next (annual survival). Through this new work, and also by making use of the legacy of records and the banding effort that has gone on at these sites for over 50 years, it is possible to look back into several decades of history in these albatross populations.

From the beginning of ornithological studies at The Snares and Campbell Island, a great effort has gone into banding both adult and young albatross at the breeding sites. Researchers such as Lance Richdale, Bernard Stonehouse and John Warham at The Snares; Jack Sorensen and Kaj Westerskov, and many Meteorological Service workers at Campbell Island; Wildlife Service and DoC workers at both sites, have contributed a major effort to build up a database of study birds.

In contrast to the foraging aspects of the studies, all that is required for recording population parameters are long hours in the field, a note-book and pencil. Over the next two years, however, analysis of the population information will be continuing in the laboratory, using computer programmes and modeling. The results of the field study are analyzed using French-developed software, which allows researchers to calculate the probability of survival of birds from one year to the next. The programme takes into account the variation in effort that has gone into searching for banded birds in different years (highly relevant when data from several decades, and different research teams are used), and also includes information about birds which were missed in a particular year, but are known to be alive from later records. Using this sophisticated approach to analyzing population dynamics, the survival of study birds can be followed with great accuracy.

This data, analyzed in relation to bycatch and fisheries effort data, along with data about the ocean, food sources, and weather patterns, will enable a thorough and objective investigation as to which factors -- such as environment- and/or fisheries-related -- are most important in affecting the populations.

As with all albatross, the mollymawks being studied have near-human life spans, taking 7-10 years to reach maturity, with some birds living to in excess of 50 years. One female Buller's mollymawk, banded at its nest on The Snares during February 1948, was found breeding close to its original nest site in March 1993. As these birds do not begin breeding until they are over seven years old, this bird was at least 51 years old when last seen. It is therefore essential to have a long-term approach to the studies. By using data which span several decades and maintaining a continuing programme of research, the analysis of cause and effect in population changes becomes feasible.

What to Eat and Where?

In addition to the population studies, all three species under investigation are the subject of detailed research into their feeding behaviour. Why? Without the knowledge of what birds are doing at sea, where they gather their food, how far they travel from their nest, which prey species they most rely on, it is impossible to say sensibly how these species might be affected by environmental changes or fishing activity. Although it is easiest for researchers to find and study these birds at their breeding sites, most of the major influences in their lives occur where they are hardest to study -- widely dispersed and highly mobile -- over the open oceans.

This research involves a more high-tech approach involving miniature electronic recorders, as well as the need for detailed behavioural observations. For the foraging study, devices used will include satellite transmitters, automatic nest-weighing machines for studying the growth of chicks, and electronic recorders to detect the presence of study birds at their nests.

Satellite transmitters, are fitted to several study birds during the breeding season. The recorders are small enough to minimize any effect on the bird's survival or behaviour. Their movements around the ocean are recorded by ARGOS satellites passing overhead. The messages are transferred to Toulouse in France, and forwarded back to NIWA by electronic mail or posted on computer disk. Although NIWA doesn't have electronic links to The Snares or Campbell Island, it is possible to radio through the locations of birds, sometimes 4-5 hours after they have been recorded, to help the field team retrieve the transmitters from incoming satellite-telemetered birds.

These studies are among the first medium-term, integrated studies of feeding ecology and population dynamics to address the conservation issues for New Zealand albatross species. We don't know whether there are currently population increases or decreases for many of the 13 taxa of albatross which nest in the New Zealand region. There is still a long way to go before we understand fully the causes of population changes. These studies will provide valuable insights into how to best protect the outstanding albatross fauna of New Zealand.

Paul Sagar is with NIWA in Christchurch
Susan Waugh is with NIWA in Christchurch.