To Predict a Bird's Song, Head Out to Space

A new study is looking at bird song evolution from outer space.

Birds come in an outstanding diversity of forms and colors, surpassed only by the great diversity of their singing. Research on bird behavior is usually focused on individuals living on a specific habitat, followed by one-to-one comparisons between different habitats. This is, for example, how we know that forest birds sing different tunes from those of the same species living on a city. Now, an international team has taken these studies a step further, by making a giant leap into space.

A team led by Dr Thomas Smith, an ecologist from University of California, Los Angeles, USA, used satellite data, combined with traditional field studies, to predict variations in songs by the common little greenbul (Andropadus virens), a songbird found in many habitats across Africa. The study, published online in the journal Evolutionary Applications, not only shows how bird songs can vary, but demonstrates how combining satellite data with field studies can trace the evolution and variation of a species behavior. It is also the first study to ever use satellite data to track variation of earth-bound species.

Little greenbuls are found throughout equatorial Africa in many different forest habitats, making them an ideal bird for this study. Meanwhile, satellite-based sensors are now sophisticated enough to make detailed "pictures" of ecological differences in terrain. In this study, researchers recorded songs along roads in Cameroon for more than 10 years, recording more than 2,000 songs from 117 birds. They then combined this data with data from several space-based sensors that measured surface moisture, biomass, elevation, roughness of terrain, photosynthesis activity, among other measures.

The habitat description made by the satellite data explained about two-thirds of the variation in songs of the little greenbul. Variations in frequency and amplitude of song were explained largely by the type of terrain. Birds singing in rain forests were particularly unique. The team then found that they could predict bird songs in other areas of Africa they hadn't tested, including Equatorial Guinea, Gabon, Republic of the Congo, and the Central African Republic.

Making such a close link between bird behavior and the habitat can help create more effective conservation programs, researchers suggest.