How Singing Species Share their Space

Ever wondered how two crickets communicate across a sea of deafening chirps – how one among a thousand Jiminys makes itself heard? Or how a couple o’ canaries chat when the air is already rife with the songs of half a dozen or so other species of whistling birds? Dr. Bernie Krause, a soundscape ecologist and father of the 1993-born “acoustic niche” partitioning hypothesis, has an answer to that. He tells us that as crickets, birds, frogs, and the like compete for real estate in the auditory market, each species carves out their own range of frequencies in song within their community. It takes a lot of energy to sing over someone else day in and day out, and that method isn’t too effective for communicating important information like predator warnings or mating pleas. Intrusive chatter is particularly troublesome for animals who can’t rely on sight – like for birds living under a dark canopy. There’s always a loser in a screaming scenario; perhaps an unheard bachelor loses a mate. Maybe an unsuspecting victim on the outskirts of a flock falls prey to a hungry cat because warning calls were lost amid a cacophony of song. Either way, living in a community with hugely overlapping song frequencies would make some species’ survival tough. So, animals find ways to avoid competition. Acoustic niche partitioning is a hypothesized solution that species seem to flock to. By carving out their own vocal bandwidths in a shared acoustic space, each species is heard in its local area even as all contribute their sounds simultaneously – a lot like radio stations broadcasting over different frequencies so that all can air at the same time without overlapping their content.

Back in the summer of 2011, Harvard undergraduate researcher Allison Ravenscraft tested this hypothesis on crickets and birds here on the Boston Harbor Islands and across a few other sites in greater Boston. She presented her findings at our 2011 park science symposium.

At the time, the results of her experiment showed more overlap in the frequencies of bird calls than expected. Her data mostly supported the acoustic niche hypothesis for insects, but it didn’t explain bird spacing or singing behavior very well.

Ravenscraft collected audio samples at a handful of different sites, then looked at the frequencies of the sounds she collected. For the hypothesis to hold water, she should have seen distinct frequency ranges spaced somewhat evenly apart on a graph. That would mean most species didn’t overlap their vocal bandwidths within close distance of one another. Instead, Ravenscraft found a fair amount of overlap in singing ranges during the morning, when birds were most active. Around midday, she found a little less overlap in song frequencies, and in the evening, when insects were active, she found little overlap. Species also seemed to share fewer singing frequencies as summer crept toward fall, likely due to the maturation of two insect species: the cricket and the katydid. As those species grew and filled a specific frequency range of their own, the acoustic data that Ravenscraft collected looked more varied. All that data seemed to point to the idea that the acoustic niche hypothesis only explained insect behavior, not bird behavior – puzzling.

Listen to NPS field collections of songs in the Boston Harbor Islands by clicking through locations on the map. Then, read on!

Years later, the now Dr. Ravenscraft has moved on from acoustic partitioning research in Boston to give a closer look at baby bacteria living in insect gut biomes in Arlington, TX. (It’s quite the interesting can o’ worms.) Although Ravenscraft stopped her acoustic partitioning work, her former mentor Dr. Brian Farrell never did. He’s kept his focus on the topic and updated Ravenscraft’s conclusions as well as those of his contemporaries.

Farrell heads a laboratory in the Department of Organismic and Evolutionary Biology at Harvard University. A little over a decade after Ravenscraft’s initial presentation here at the Boston Harbor Islands, Farrell and his colleagues have published new research that fills out more details in the story of the vocal patterns of our winged neighbors. In his December 2022 paper, Farrell discusses why so many studies on this topic have reported such mixed results. He writes that birds may use strategies to avoid vocal competition that aren’t considered in study designs – strategies like moving further away from other birds with similar pitches rather than changing their own song. Or, birds may simply sing at different times. Low quality microphones and study designs themselves could also be culprits in misleading researchers who are interpreting the data. If microphones pick up too large or too small a range, data won’t be representative of the sounds specific to birds in their respective space. Farrell looks at a range of about 100 meters per site to avoid this problem. He also compares singing patterns across different spaces to address other strategies that birds use to avoid competition.

The traditional method of analysis for this work is often to blame too. In Farrell’s recent research, he considers factors often left out of the standard model so often used. Mainly, among birds, body mass correlates to pitch range – the larger the bird, the lower the call that they make. Birds tend to be bigger in the tropics and smaller closer to Earth’s poles; that translates to more low-pitch songs in the tropics and more high-pitch songs in the far northern and southern reaches of the globe. Without taking into account bird size, acoustic data can look different from what we’d expect, and that can confuse interpretation. It could look like several species are overlapping their song ranges, when really, there are more birds of the same species singing in similar registers.

This is an example of a well-understood lesson in science – it can be easy as pie to draw the wrong conclusions from complex data. Even for the sharpest eyes and most rigorous study design, researchers may not know or realize that a factor is relevant until new information comes along or until underemphasized perspectives are amplified. That's part of why we constantly revise and repeat studies. It's why peer review and collaboration are so valuable.

Farrell and his team adopt a new model that considers bird mass as they listen to tropical bird songs across South America.

After factoring in the birds’ mass, Farrell finds evidence of acoustic partitioning. Big news for birding biologists!

Farrell also creatively applies his data to address other, less-tested facets of the hypothesis, strengthening its validity. He shows that in similar habitats, birdsong ranges look similar, falling into similar frequency distributions. Birds were scooting out of range of others who sang in their preferred pitches and then carving out that acoustic space for themselves in nearby areas. That’s a cool spatial feature of Krause's hypothesis. Interestingly, Farrell’s new research highlights how birds’ calls change depending upon their habitats too. Birds in tropical rainforests sing within a smaller range of frequencies than birds in tropical cloud forests. This could be because lower elevation rainforests have more insects singing at high frequency, preventing birds from using a high register. Then again, maybe rainforests have larger fruits that attract larger birds who sing in a lower register. Either way, partitioning is far more expected in tropical regions like these, where bird populations are denser, rather than in spaces like the Boston Harbor Islands, where birds have more of an option to spread their wings and skitter to another space to sing without competition.

In their paper, Farrell and his team still write that other considerations, like species abundance and mass, ought to be more carefully incorporated into future studies to better understand just how significant the competition for acoustic space is in shaping bird communities. And beyond exploring the influence of these factors, still more questions linger in the air. Could bird song partitioning be a consequence of evolution rather than behavior? Or could a birds’ mass be the key influencer for song rather than merely a biasing factor? Farrell and his contemporaries ruminate on these among other questions.

For now, however, Farrell and his team have demonstrated that birds choose their homes, at least to some degree, based upon what their neighbors sound like – not so different from us! When you opt to rent a home in suburbia rather than an apartment above a rave warehouse, you’re thinking a bit like a bird…and a cricket!