Tracking the Twitter bird

Kevin Fraser, University of Manitoba

The Twitter logo is ubiquitous and familiar to millions of users, but the bird it is based on is relatively little known. Let me introduce you to the Mountain Bluebird, a species that Henry David Thoreau once noted, “carries the sky on his back”. This stunning bird breeds in the grasslands of Canada during our summer before migrating southward to spend our winter in the southwestern U.S. and northern Mexico.


Above left, Mountain Bluebird perched on a fencepost near his nest in central Alberta, Canada (photo Myrna Pearman) and right, this silhouette needs no introduction, but is based on Mountain Bluebird.

But while the Twitter bird population continues to grow and proliferate with every Tweet, the Mountain Bluebird is steadily declining for mysterious reasons across much of its range. Birds that inhabit grasslands are suffering some of the steepest population declines of any group of bird (State of Canada’s Birds 2012). One major barrier for understanding and mitigating these declines is that we do not know exactly where birds go after they stop tweeting at their breeding sites.

Enter new technology that allows us to track the migration of small birds. Some Mountain Bluebirds in Alberta are now carrying both the colour of the sky on their backs as well as tiny, electronic devices that will track their migration all the way to their wintering sites and back. This summer with collaborators Biologist Myrna Pearman from the Ellis Bird Farm, Citizen Scientist Brian Biggs, and others, I began an inaugural tracking project to figure out just where Mountain Bluebird spend our Canadian winter, and how they get there. What we learn about migratory connectivity and important habitat used by bluebirds year round may give us something to Tweet about when the birds return next spring.

Mike Symington of Calgary’s CBC joined us for field work in Alberta and produced a video – view it here.


A Mountain Bluebird outfitted with a light-level geolocator that will track his migration to the southern U.S. or Mexico, and back (photo Kevin Fraser).

What happens when migratory songbirds are not on time?

By Kevin Fraser

Spring migration is like an ocean tide, flooding back northward until woodlands, parks and backyards are once again filled with the exuberant, liquid singing of migratory songbirds. The timing of migration also has a satisfyingly predictable ebb and flow; the countless trials and errors of individual birds over centuries are visible in the patterns we see today. If their timing is right, songbirds can avoid the harshest spring weather. They can get good places to nest, find their choice of mates, and enjoy the bounty of spring insects that they need to be successful parents. But new weather extremes and climate change may be disrupting these ancient patterns, causing birds to get out of sync with their environment.


Figure 1. Purple martin perched at a North American breeding colony. Martins migrate 10 000-20 000 km annually between breeding sites and overwintering roosts in South America (photo by Kevin Fraser).

Long-distance migratory songbirds, particularly those that migrate the furthest and feed on aerial insects (aerial insectivores), are showing steep population declines (State of Canada’s Birds 2012). There are likely many factors working in tandem to cause such declines, but could there be anything big and broad enough to affect all of these species at once?

One likely culprit is climate change and new extremes of weather, which may be affecting the relationship between migratory songbirds and their food. Research with the Pied Flycatcher (Ficedula hypoleuca), a species that migrates between Europe and Africa, has revealed disturbing trends: as springs have become warmer over decades in Europe, the caterpillars preferred by Pied Flycatcher are available much earlier than they used to be. However, Pied Flycatchers seem to be missing the new dinner bell. Their migration timing has not changed much over the same time period, putting their arrival at breeding sites out of sync with peak caterpillar availability (Both et al. 2001). It is as if a dinner party’s time has been changed without your knowing, and when you finally arrive all the hors d’oeuvres have been eaten and everyone else is well through the main course. But, in the case of the Pied Flycatcher, the consequences of this mistiming are severe; they are declining the most sharply in locations where caterpillar peaks are the earliest (Both et al. 2006). These late arriving birds miss most of the food and are therefore less able to feed and fledge lots of young, which eventually results in population declines overall.


Purple Martin male w geotag, 2014-02-24 (1)

Figure 2. Purple Martin equipped with a light-level geolocator (visible on lower back) that will track migration to South America and back (photo by Jerome Jackson).

Much of my Liber Ero-supported research is focused on the Purple Martin (Progne subis) – an aerial insectivore that migrates between North and South America (Figure 1). On a decadal scale, we know that Purple Martin have been arriving earlier at their breeding sites as springs have been getting earlier (Zelt et al. 2012). But are martins changing their timing everywhere and are they doing it fast enough? Do individual martins have the flexibility they need to make adjustments in their year-to-year timing in response to changing climate and weather? Now that we have new tracking technology (geolocators) to follow the migration of individual birds (Figure 2), we have a unique opportunity to investigate how flexible martins are to different weather conditions they experience while migrating in spring. For example, did Purple Martin hear the early dinner bell in 2012, when spring was the earliest and hottest on record in eastern North America? If so, did they adjust their migration timing? We learned that the disappointing and short answer, is, no. We found that in 2012, martins did not depart earlier on spring migration from the Amazon, nor did they travel any faster than in previous years. They therefore did not arrive earlier at their breeding sites to match the advanced spring of 2012 (Fraser et al. 2013).

Part of the issue may be that martins are not receiving the weather signals they need to be able to adjust their migration timing successfully. To better understand what a martin on migration experiences, we measured weather conditions all along migratory routes, from the Amazon all the way up to their breeding sites. This is something we could not do until geolocators enabled us to identify the specific routes and timing of individuals. We found that temperatures were similar to previous years along much of the migratory route, and there was no indication of an earlier, hotter spring in 2012 until martins were back in North America and almost at their breeding colonies, thus they had little opportunity to hit the accelerator and arrive earlier (Fraser et al. 2013). So for martins, the insidious effects of mistiming could be contributing to population declines.

What can we expect of Purple Martins this spring with one of the latest and coldest springs on record? Early reports note that breeding arrival dates are more than a month later in some areas this year. It may be that martins are able, or forced, to put on the brakes during cold weather but are unable to hit the accelerator with an advanced spring, as we found in 2012. The data we are gathering with collaborators across the range will help us to answer these questions and more, so stay tuned.

We are now continuing to track martins with geolocators to measure directly how migration timing affects their ability to successfully raise young. This may not be the same each year and could vary widely by breeding location. Do the relative advantages of early or late migration change each year depending upon breeding location and the weather? Perhaps the wild variation in spring weather we have been experiencing in recent years will have bigger effects in some locations than others, which may help to explain the steeper population declines we see at the northern end of the breeding range for many aerial insectivores (Nebel et al. 2010, Fraser et al. 2012). We need to know these patterns so that we can better understand what is causing stronger declines in some places and prioritize effective conservation actions.

My Liber Ero research is conducted in partnership with many organizations, including the Purple Martin Conservation Association ( An extended version of this blogpost appears in the spring edition of their Purple Martin Update Magazine.