Testing the krill surplus hypothesis

Exploitation of living resources in the Southern Ocean potentially has a profound impact on the ecosystem. Perhaps the greatest impact comes from the harvesting of large krill-consuming whales, whose populations declined dramatically during the 20th century as a result of whaling. Twenty-five years ago it was proposed that this reduction in whale populations led to a ‘surplus’ of uneaten krill – up to 150 million tonnes per annum – which became available to other krill consumers. This proposal became known as the ‘krill surplus hypothesis’ and strongly influenced thinking on the krill-based ecosystem in the Southern Ocean.

One prediction of the krill surplus hypothesis has been that the excess krill has allowed other populations of krill consumers to increase. The crabeater seal, which eats krill rather than crabs, has been the subject of much speculation about current populations and change in populations over the past 50 years. Crabeater seals are restricted to the pack-ice surrounding Antarctica and have a circumpolar distribution.

The earliest surveys in the 1950s estimated the circumpolar crabeater seal population to be around 5 to 8 million animals. Later surveys in the 1970s produced estimates of around 12 million animals. This apparent increase in population size was taken as evidence in support of the krill surplus hypothesis. In the early 1980s it was predicted that crabeater seal populations would increase to 50 million animals by 2000 – a 10-fold increase over the last 50 years (and a four-fold increase over the past 30 years). If true, this would represent a fundamental change in the structure of the krill-based ecosystem.

This prediction has fuelled speculation about the dominance of the crabeater seal as a consumer of krill, over other animal groups such as birds, fish and squid, and led to claims that the crabeater seal is the most ‘abundant large wild mammal on earth’. It has even been suggested that crabeater seal populations could hinder the recovery of exploited whale populations because the krill surplus is no longer available to the whales.

Until now it has not been possible to substantiate or refute the predictions of the krill surplus hypothesis because there have been no estimates of crabeater seal populations since the 1970s. A recent Antarctic Pack-Ice Seal program (APIS) aimed to address this situation.

APIS was an international, multi-platform survey of the regional and circumpolar abundance of pack-ice seals. Australia took a lead role by undertaking a major survey off east Antarctica in the summer of 1999-2000. This survey, which extended from 60°E to 150°E, straddled most of the coastline of the Australian Antarctic Territory and covered one quarter of the longitudinal extent of circumpolar pack-ice (Australian Antarctic Magazine 1: 7-8). The boundaries of the survey region coincided with surveys undertaken in the 1970s (when the crabeater population was estimated at 800 000), allowing a comparison of regional crabeater seal abundance in the same area and at the times (1970s and 2000) of the krill surplus hypothesis predictions.

Estimating the abundance of any animal over large, remote areas is always a challenging task, but there could be few more difficult surveys than for crabeater seals in the pack-ice. In the 1999-2000 summer, observers used the Aurora Australis and two Sikorsky S76 helicopters to count seals distributed across 1.5 million km2 of pack-ice, in strips either side of survey tracks. As there was no guarantee that an observer would be able to count all seals on the ice in the strip, it was difficult to estimate the number of seals present but missed.

Another difficulty was estimating the number of seals in the water when the ship or aircraft passed by. This required capturing some seals and deploying dive recorders to record the time they spent hauled out and on the ice.

The next problem involved ‘scaling up’ counts from the survey strips to the entire survey region. Traditional scaling up methods require the survey tracks to be scattered at random through the pack-ice. However, this was not possible because the ship was limited in its movements through the ice and the helicopters were limited by weather.

After developing new methods of data collection and analysis to address these problems, the ‘best estimate’ for the survey region was around one million crabeater seals, although plausible estimates ranged from 700 000 to 1.4 million.

These logistical and estimation difficulties were also present in the 1970s survey, but many of the technical advances that could be applied in the 1999-2000 survey were not available 30 years ago. The 1970s estimate of around 800 000 crabeater seals may therefore be biased. Even if we allow for bias, there is no strong evidence in support of the four-fold increase from the 1970s to 2000 that is predicted by the krill surplus hypothesis.

And what of the predicted circumpolar population of 50 million crabeater seals in 2000? Testing this prediction must await the analysis of data collected by the other nations who participated in APIS. However, with only one million or so seals estimated to be in a zone covering one-quarter of the circumpolar region, it seems hard to imagine another 49 million in the remaining three-quarters!

After 25 years of discussion and speculation about the possible impact of a krill surplus, there is finally some evidence to examine whether the predictions can be substantiated. The results of the Australian APIS survey suggest that we may need to re-assess the krill surplus hypothesis and, more broadly, our understanding of the structure of the krill-based ecosystem in the Southern Ocean.

Improved modelling of the krill-based ecosystem is now a priority activity for scientists working within both the International Whaling Commission and the Convention for the Conservation of Antarctic Marine Resources. The results from the APIS surveys will be pivotal to these modelling efforts.

COLIN SOUTHWELL, Southern Ocean Ecosystems Programme, AAD