Scientists drop in on little-known, rarely sighted beaked whales

A Gray's beaked whale, seen near Whangaparaoa in 2009. There are about 22 species of beaked whales globally, of which about 13 have been found here, but little is known of them. Photo / Brett Phibbs

Scientists who placed high-tech listening instruments in the Cook Strait believe they've picked up the sounds of a little-understood and rarely sighted species of whale.

Niwa scientists confirmed the presence of the yet-to-be identified beaked whales by analysing underwater acoustic data collected during two six-month deployments of six passive acoustic moorings.

In the first project of its kind in New Zealand waters, the moorings recorded the entire underwater soundscape of the region, including sounds produced by marine mammals.

One of the aims of the project, just featured in the latest issue of the Journal of the Acoustical Society of America, was to learn more about the presence and distribution of whales and dolphins in the region.

But the two types of signals from beaked whales were identified in the Cook Strait region differed from any beaked whale signals previously documented.

There were about 22 species of beaked whales globally, of which about 13 had been found in New Zealand, but very little is known about them.

Sightings are rare because they are deep-diving animals that can spend more than an hour on a single dive and surface for a very short time.

"Most information about their presence in New Zealand waters has come from people reporting whale strandings but it is very limited," said Niwa marine mammal expert Dr Giacomo Giorli, who carried out the work with colleague Dr Kim Goetz and JASCO Applied Sciences experts.

"We really know very little about their behaviour."

What was known is that beaked whales, like all toothed whales, such as dolphins, sperm and killer whales, used echolocation, emitting sounds and listening to the echoes, to locate prey in the dark deep-sea environment.

The new signals were extracted using a multi-step process.

First, an automated detector was applied, which was then validated through manual analysis.

This process involved researchers examining a subset of the detections by looking at spectrograms - visual pictures of sound frequencies – to verify that the detections were correct.

A click-by-click analysis was then undertaken.

From the reviewed subset of the data, which comprised just 1 per cent of the total collected, three distinct beaked whale signals were detected.

One matched the Cuvier beaked whale – one of the most commonly seen beaked whales – but the other two did not match those from any previously recorded beaked whale species.

Giorli speculated that the most likely possibility, based on stranding records, is that the signals belong to the Gray's and strap-toothed beaked whales.

But matching the recordings of the beaked whales to the correct species could only be done by concurrent visual and acoustic observation, which had not happened yet.

The next stage was to examine the rest of the data set for similar signals to determine how often the sounds from the unidentified beaked whales were recorded.