Sunday, November 29, 2009

Birdsongs provide population clues

Microphone array improves estimate of Ovenbird density.
Emma Marris
Published online 27 November 2009 Nature doi:10.1038/news.2009.1112

As any birder will tell you, most birds in the forest are easier heard than seen. Now two scientists have figured out a way to estimate bird population densities by recording their songs with an array of microphones.

The method offers an alternative to a common way to estimate population densities: the human ear. Humans listeners are often used in bird studies, but people are far from perfect, says Murray Efford of the University of Otago in Dunedin, New Zealand. In particular, "we aren't good at telling how far away sounds are," he says.

Efford and Deanna Dawson of the United States Geological Survey in Laurel, Maryland, have come up with a method that uses multiple microphones scattered through the woods. By recording in several places simultaneously, researchers can estimate each bird's acoustic 'footprint' — the area around it where it can be heard.

The size of the footprint depends on parameters such as the loudness of the birds and the acoustic properties of the forest. So Efford and Dawson must try different values for such parameters until they find a good match with the data recorded by the microphones. When all is done, the duo can estimate bird density without knowing the birds' locations or the size of the forest.

Twitter tweet!

They tried out their method on ovenbirds (Seiurus aurocapilla) at the Patuxent Research Refuge near Laurel, Maryland. Only male ovenbirds sing, and the technique estimated their density at around one male bird per every five hectares (click here to hear the ovenbird's song). The findings matched well with estimates gleaned from catching the little songbirds in nets. What's more, the researchers found that the new technique was more precise than estimates based on netting. The work is published online in the Journal of Applied Ecology1.

The researchers say that the method could be used to estimate densities of other hard-to-spot animals, including whales and dolphins. Len Thomas, a statistical ecologist at the University of St Andrews, UK, for instance, is already using a similar method as part of an effort to monitor Minke whales (Balaenoptera acutorostrata) by their sounds. Sightings of these whales in the Pacific can be counted on one hand, but males make a distinctive "boy-yoy-yoing" sound, so hydrophones can measure their song footprints, just like the ovenbirds.

However, Thomas says that Efford and Dawson's method only provides part of the picture for Minke populations. The method estimates only the densities of sounds, not of animals, and in the case of the whales, uncertainty about what percentage of males call and how often they do so make it hard to extrapolate to an estimation of the population density.

Efford adds that the new technique will work best with animals that make repetitive sounds at constant loudness. That means it could be especially useful for estimating population densities of other kinds of birds. "A lot of birds are blurting out the same thing over and over again, persistently and monotonously," he says.

The monotony may have gotten to Efford, after listening to ovenbird songs over and over for the study. "It is a particularly irritating and insistent call," he admits.

References
Dawson, D. K. & Efford, M. G. J. Appl. Ecol. 46, 1201-1209 (2009).