As part of my series of interviews with women in science, in this post I will be interviewing Carrie Branch. Carrie Branch is a PhD candidate at the University of Nevada, Reno, working with Vladimir Pravosudov. In a recent study, Carrie found that chickadees sing different songs at low and high elevations and between populations on different mountains. Here I ask her some questions about what these results mean.


First, what are mountain chickadees and why do you study them?

Mountain chickadees are small songbirds that inhabit the montane regions of western North America. These birds cache, or store, food in the summer and fall to use later in the winter when food is scarce. We know that they use spatial memory to relocate their caches, which makes them interesting for asking questions about cognitive ecology.

In addition, because they inhabit a continuous elevation gradient in the mountains, different groups of birds experience differential pressure on their ability to find their food caches. For instance, birds at higher elevations experience colder winters, with more snow and longer snow cover compared to birds living at lower elevations. As such, the birds at higher elevations may need to rely more heavily on their spatial memory ability to relocate their food stores. And this is exactly what work in our lab has found, birds from higher elevations cache more food items, have better spatial memory ability, and have a larger hippocampus (the brain region associated with spatial memory) with more neurogenesis, than birds living at lower elevations. Furthermore, my lab mate, Dovid Kozlovsky, and I showed that low elevation birds explore a novel environment faster and are socially dominant compared to high elevation birds. All of this work taken together sets the stage for asking really interesting questions about how these birds, who are the same species, differ across various cognitive, behavioural, physiological, and morphological characteristics as a result of climatic variation.

A mountain chickadee. Credit: Carrie Branch

Why were you interested to know if chickadee song varied between populations on different mountains and between different elevations on a single mountain?

Based on the previous work our lab has conducted, we hypothesized that there may be separation between high and low elevation birds, as high elevation birds would likely have lower fitness at low elevations because of their socially subordinate status, and low elevation birds would likely have lower fitness at high elevations because of their poor spatial memory ability.

This prompted me to ask whether female mountain chickadees contribute to this hypothesized separation between high and low elevation birds by preferentially mating with males from their respective elevations.

Song in temperate male songbirds has two functions: it is used to defend territories against other conspecific males and is used to attract females. Song is learned from the male’s social male parent, so it can vary geographically, potentially creating dialects between groups of birds that are the same species but separated by geographic barriers or local adaptations.  As such, female songbirds can use song as an indicator of a male’s location of origin. Females prefer song from local males, because local males likely have the proper phenotypes to be successful at their location, compared to males from another location (particularly if local adaptation is at play). 

I then asked whether song structure differed between males from high versus low elevations. Male song is plastic, and can be modified in response to varying auditory environments. One major hypothesis for why song varies geographically is the acoustic adaptation hypothesis, which suggests that songs differ because birds alter their songs to maximize transmission in a given habitat. As such, we needed to record and assess song structure at another mountain location with similar elevations and forest composition to test whether differences in song structure were consistent with identifying a male who is locally adapted or if the male’s song had simply been altered to fit the environment. Our finding that song structure differed at all four locations, despite similar forest composition at both high and both low locations, is inconsistent with the acoustic adaptation hypothesis, and instead supports the indicator of locally adapted males hypothesis.

Carrie Branch with a chickadee. Credit: Carrie Branch

How do you go about testing this in a chickadee? What is it like working with birds in fairly extreme montane environments?

I recorded male song during the spring, when males primarily sing. This simply involved listening for a male to sing and then approaching the male and recording the song using a digital recorder and microphone. Males were not all individually identifiable so I made sure to leave at least 500 m between each recording. I then uploaded the recordings onto the computer and used SIGNAL bioacoustics sound analysis software to extract 19 measures from each song, around 6-30 songs from each male recorded. I used the average of those measures as a representative for one bird to avoid pseudoreplication. These acoustic measures include the start and end frequency of each note in a song, as well as the duration of the song and each note, time between each note, and others.

Working in the mountains is quite difficult in the winter when there is a lot of snow. We often struggle to get to our field sites and have to use snow machines to do so. The snow typically melts around May/June, but I have to record birds starting in April/May so I often walk around in the snow and just listen for birds to sing. If the snow is not too deep its really not a problem, but sometimes it is difficult to follow birds, as they are highly mobile. I often use playback recordings of mountain chickadee song to entice males to approach me and sing.

What can your finding tell us about songbird song and cognition more generally?

I think what is really interesting about our findings in this study are that we see song structure differentiation on a really small spatial scale with no geographic barrier preventing movement between high and low elevation birds. This of course leads to the question of whether or not these structural differences represent true dialects or if these differences are simply consistent with clinal variation. I have now recorded song along the elevation gradient at both mountain locations so we can assess this question. If these structural differences represent true dialects, we should see a break in the acoustic measures where birds below a particular elevation, say 2100 m, where rain changes to snow in the Sierra Nevada, sing the “low” dialect and those above 2100 m sing the “high” dialect. Preliminary analyses suggest we may have evidence for dialects in these two mountain locations.

Do you have any advice for other researchers who might want to study cognition in the wild?

Do it! We have answered a lot of questions in the lab, and there are obviously still many more, but assessing cognition in the wild allows us the opportunity to investigate real behaviours that are important to animals and, I think we get truer answers to our questions. Lab settings are great for controlling variables, but often times animals are very stressed and asked ecologically irrelevant questions. It is so important to bring our questions back to the animals’ ecology in their natural environments where we can hopefully capture true variation. Persistence is key, as field work can be rather messy.


Reference: Branch, C. L., & Pravosudov, V. V. (2015). Mountain chickadees from different elevations sing different songs: acoustic adaptation, temporal drift or signal of local adaptation?. Royal Society open science2(4), 150019. DOI: 10.1098/rsos.150019