Movement Ecology of Animals

Homepage of Dr. Emily A. McKinnon


Connecticut Warblers fly over the Atlantic for 2 straight days in the fall

Yep. Move aside, Blackpoll Warblers. There’s a new trans-Atlantic fall migrant in town. Okay, actually Connecticut Warblers have probably been doing this for thousands of years, but it’s pretty surprising because we didn’t know it until now.

In my Scientific Naturalist paper in Ecology (or download directly McKinnon_et_al-2017-Ecology), I show how 4 Connecticut Warblers flew from the coast of the US for two solid days (minimum of 48 hours) over water to land somewhere in the Greater Antillean islands, probably Hispaniola. To make it even more stunning, they rested for about a week, then again flew almost 800 km over open water to the Gulf of Venezuela. And they weren’t done yet! They continued south to winter somewhere deep in the Amazon basin. The data are a little blurry on the exact wintering sites, but somewhere deep in the green that is the Amazon forest.

In contrast, the ocean-crossing is as clear as day because, well, there are no trees over the ocean! So when we are using light-sensors to track migration, it’s really obvious when the birds are not experiencing any shading at all from trees or other vegetation. The light levels just steadily increase and then decrease as dusk falls. In contrast, when the birds are on land, it’s a mess of light-levels going up and down, even as low as zero when birds are in really shady spots (giving us dozens of artificial ‘sunsets’ that we have to filter out). This is partly to blame for our fuzzy delineation of the exact overwintering sites for these individuals. The Amazon forest understorey is a dark place!


Connecticut Warbler with a light-sensing geolocator backpack. This was one we put out this past summer (2016) so we hope this bird comes back this year (2017)! The stalks will hopefully give us better resolution for identifying wintering sites. The original tags we used were stalk-less and had a lot of shading from the feathers.

Back to the fall migration of this species – how did we miss this amazing feat of migration?

I think there are a lot of reasons, actually. First of all, Connecticut Warblers, on a global scale, are not very abundant, so they are not a species commonly captured or seen by banders or birders. They nest in the southern-boreal aspen-transition habitats in Manitoba, Ontario, Quebec, and some parts of the northern Great Lakes states like Wisconsin and Michigan, but they are not common even within their breeding range. Not many people venture out into bogs in June in Manitoba either, and I know why!


Feeding the healthy mosquito population while catching Connecticut Warblers

Okay, so not a lot of overlap between people and Connecticut Warblers (a range map is in my previous post on these guys). I also think that Connecticuts probably get confused with their more southern-breeding cousins, the Mourning Warblers, or their western cousins, MacGillivray’s Warblers. This would be especially the case in the fall, where the young-of-the-year in those species might actually have a complete eye ring like a Connecticut Warbler (normally they only have semi-circles, as in MacGillivray’s, or none, like the Mourning).

So all of this means that coming across one of these guys and positively identifying it in the fall especially might be tricky.

But we had some hints that they might do something drastic over the ocean. Hurricane Emily brought down 75 individuals on BERMUDA one October. What would they be doing way the heck over the ocean?!? There is also a really neat video capture of a flock following cruise ship lights in the Lesser Antilles. Finally, the famous ornithologist Alexander Chapman commented on how darn FAT Connecticut Warblers were in the fall in the northeastern states… hmmm, suspicious, no?

Now that we have some tracking data, I can see additional reasons why we would miss this. First of all, if their major Caribbean stopover is on Hispaniola, it would be easy to miss. Hispaniola’s landscape is rugged, and humanitarian crises have resulted in few bird surveys there in recent decades. Unfortunately, it’s also lost a lot of forest recently- something like 90% of the original forest cover in Haiti is gone! Yikes.

Now that we know Connecticut Warblers travel through this area, hopefully we can get some more details on what type of habitat they might use or need for these Caribbean stopovers. Clearly their strategy relies on extraordinarily long migratory flights, which suggests they also need extraordinary amounts of fuel in advance of these flights as well as for recuperation after.

For Connecticut Warblers, I hope this study gives them some well-deserved time in the spotlight. They’ve been missing in action for too long, and I don’t want them to end up missing in action for good!

Full citation for the paper plus links to download a pdf and supplemental files:

McKinnon, EA, Artuso, C., and OP Love. 2017. The mystery of the missing warbler. Ecology. 10.1002/ecy.1844


ecy1844-sup-0001-AppendixS1 (1)



Connecticut Warblers of Manitoba.




Leave a comment

Would you like a side of fruit with your beetles and ants?

One of the interesting things about Wood Thrushes and many other migratory species is that they have an amazing plasticity in diet. Baby songbirds grow up getting half-smooshed arthropods shoved down their throats from Mommy and Daddy. The high-protein and high-fat content of insects and spiders make these a good food source for growing nestlings. There are some exceptions to this, but in general, baby-bird food is animal-based. Adult songbirds have a lot of different strategies. The thrushes and many warblers also eat a lot of ‘bugs’ during the breeding season, when summer productivity is high and there are lots of juicy lepidopteran larvae around. In fall though, many switch to gorging on fruits. Many songbirds are key seed-dispersers actually, and the many red-berried fruits decorating fall forests are a testament to the co-evolution of birds and plants (see the Wood Thrush-American Ginseng example here).

During the overwintering period, many songbirds continue to eat fruit, and some even take on a nectarivorous diet, like the Cape May Warbler Setophaga tigrina (Latta and Faaborg 2002). When birds do eat a mixed diet of animal and plant-based foods, is there an optimal combination for staying healthy? Or are plant foods just a poor-quality place holder until more arthropods can be found? This is a question that has been looked at in a few species of overwintering migratory songbirds, but with mixed results. Hermit Thrushes (Catharus guttatus) seem to fatten up more easily on diet of arthropods (Long and Stouffer 2003). But in Costa Rica, migrant songbirds as a group seem to increase their consumption of fruits later in the nonbreeding period, a time when they should be starting to fatten up and pack on muscle for migration (Blake and Loiselle 1992).  This got me wondering about Wood Thrushes. I found out that the forests at my study site in Belize dry out seasonally, and the abundance of arthropods declines. Maybe Wood Thrushes would switch to a more fruit-filled diet later in the winter, prior to migration. If so, would there be a cost? Could they still get ready for their spring sprint northwards, feeding on tropical figs instead of ants and beetles?


Most awesome Wood Thrush photo ever by Chris Jimenez. All his photos are fabulous but this one is particularly impressive because 1) it’s actually in a tropical forest (in Costa Rica), 2) it’s eating a beetle, and 3) Wood Thrushes are hard enough to see in dense tropical understory, let alone photograph!  I really wanted to publish this paper in a journal with a cover photo so I could use this picture, but alas Journal of Field Ornithology doesn’t do cover photos. 

[Check out more of Chris’ amazing photos here:

To figure this out I analyzed the diet of Wood Thrushes from my study site in Belize. It can be challenging to figure out the diet of small songbirds, since they are not easy to watch in the wild. I don’t know how many times I saw a Wood Thrush foraging on a trail, then ran over as soon as it flushed to try to see what it had been eating. Usually there was no sign of anything! It was more obvious when they were eating fruit, since they would join up with dozens of other Wood Thrushes and many other species, like the tropical resident Clay-coloured Robins, Black-faced Tanagers, Gray Catbirds, and even toucans like the Collared Araçari, all in the tops of fruiting trees, generally making an obvious ruckus and spitting out seeds all over the ground under the trees. I saw this most frequently with Ramon or breadnut trees – Wood Thrushes and other birds loved these fruits, and I could catch Wood Thrushes all day in a single net at the base of the tree, as birds from all over came in for the bonanza.


This is a Wood Thrush trying to eat a ramon fruit. These fruits are 90% seed, with a thin coating of fruit around the outside. Wood Thrushes spit out the seeds on the spot. Clearly they are not doing a very good seed-dispersal job. Maybe the bigger birds like toucans are what this tree is really hoping for? 

To analyze the diet of Wood Thrushes, I decided to look at the stable isotope ratios of nitrogen and carbon in their blood. Basically, there are multiple isotopes of these common elements, and animals preferentially incorporate the lighter isotope into their tissues. This means that animals eating things higher up on the food chain will be consuming tissues with more of the heavier isotope. For example, a top predator, like a jaguar, would have a higher stable isotope ratio of Nitrogen compared to a herbivore like a tapir. Carbon works basically the same way, but also with some differences in the isotope ratio of different types of plants (e.g. grasses versus herbaceous). These isotopes are useful for distinguishing a diet of plant- vs. animal foods because the animals should always have higher stable nitrogen isotope ratios (and to a lesser extent carbon ratios) than the plants. That means a Wood Thrush eating more bugs will also have a higher stable nitrogen isotope ratio in its tissues. For this project, I was already collecting a tiny blood sample for genetic sexing so I used the remainder of the blood for the stable isotope analysis.


Taking a tiny blood sample for genetic sexing and stable isotope analysis.

I also collected a bunch of potential food sources for comparison – beetles, ants, fruits, spiders, and grasshoppers. First I just looked at the isotope ratios of carbon and nitrogen, then I tried to reconstruct the actual proportions of different foods in the diet by using a mixing model.

Here’s the basic food web that I reconstructed:


On the vertical axis is the ratio of stable nitrogen isotopes, which we call delta-15-N, measured in units of permil (that weird percent thing in the brackets). As you go up the food chain, delta-15-N gets higher, which is why spiders and Wood Thrushes are in the top part of the graph, and fruits and grasshoppers at the bottom. The horizontal axis shows the stable carbon isotopes, or delta-13-C. This plot is call a C-N plot. The dots are the average values, and the crosses show the variation around the average. You can see that Wood Thrushes are all the same (no visible error bars) but beetles and spiders are really variable, especially in delta 15-N. This is probably because I lumped a lot of different species of these guys together. 

Right away, the data seemed to suggest that Wood Thrushes were not eating that much fruit. Their stable nitrogen ratios were just too high! I analyzed the data to see if the highest stable isotope values were associated with the best body condition in Wood Thrushes. Did birds at the ‘top’ end of the food web have a payoff compared with birds slightly lower down? The answer is a resounding no, although, there wasn’t a lot of variability in the diets of the Wood Thrushes I analyzed. Most of them seemed to be eating primarily arthropods, and the slight variability across individuals didn’t predict their body condition, fat or muscle levels. But there were some interesting patterns in the data that were somewhat unexpected.

First, I did find some differences by habitat. Birds in the driest, early successional forest tended to eat more fruit mid-winter but then actually increased the amount of arthropods they consumed as spring approached. The opposite pattern was found in the more mature forest sites, with the end result that diets were actually very similar across all habitats in late winter. This tells me that Wood Thrushes in different habitats might be fine eating whatever is easiest and available all winter, but as spring approaches they seek out an optimal balance of fruit and arthropods. This habitat difference also likely explains a few differences I found between males and females, since females were more abundant in the dry habitat. The diet differences didn’t relate to body condition though, so it wasn’t that females or birds in the dry habitat were suffering because of their different diets. They just ate different stuff.

So seems like bugs with a side of fruit is what Wood Thrushes go for in the winter (and not fruit with a side of bugs). What exact types of bugs and fruits would be interesting to figure out – maybe something like DNA barcoding will give us more specific information on their diet in future.

Check out the full paper here:

McKinnon, EA, TK Kyser and BJM Stutchbury. 2017. Does the proportion of arthropods versus fruit in the diet influence overwintering condition of an omnivorous songbird? Journal of Field Ornithology. Early online. doi:10.1111/jofo.12187




1 Comment

The most interesting boreal bird you’ve never heard of

Spring is on the horizon and birders everywhere are starting to get antsy for those first early migrants. This year is a very exciting one for me, as I’m waiting for the return of one of my newest study species – one of the most mysterious boreal songbirds out there. Right now, ‘my’ birds are out there with tiny data-logging backpacks on, fattening up deep in the jungles of South America, waiting to head north.

Let’s see if you can guess who they are. These small warblers are long-legged, but not always found on the ground; they are colourful, but camouflage well; they are loud songsters, but largely quiet on migration, and often missed at breeding sites; they breed in open bogs, mature aspen forest, or mature mixed woods, but most people live far away from their breeding range; they look very similar to two other warbler species, but they are genetically distinct so placed in their own genus. Here’s a recording I made of one this past spring if you need a big hint:

The answer is… Connecticut Warbler (Oporornis agilis)!

First let me just explain that this unfortunately-named species has little or nothing to do with Connecticut, at any point in its life cycle! The name came from a single specimen that was ‘collected’ (i.e. shot) over the state of Connecticut in 1812 by Alexander Wilson. A better name for this species might be something more physically descriptive, like blue-headed or eye-ringed warbler, or maybe something related to habitat, like boreal bog warbler.


Check out that beautiful eye ring! Thanks to my field tech @KelseyDBell for taking the photo

In general Connecticut Warblers have a reputation for being mysterious, shy, retiring, skulky, rare, or even ‘sluggish’! After my first field season with this species this past summer, I think this reputation is because most birders only ever see Connecticuts during migration, and they are fairly rare and quiet during this period. Connecticut Warblers also migrate later in spring than a lot of other species, so they easily miss peak spring birding in May. Fortunately for me, living in Manitoba, at the edge of the boreal forest, Connecticut Warblers can be found breeding within 45 minutes’ drive of my house! Delving into the scant literature on this species last winter, I realized that we might actually be losing the Connecticut Warbler before we even know anything about it.

Long-distance migratory songbirds as a group are showing population declines across North America – my previous study species, the Wood Thrush, shows a fairly typical pattern of about 2% population loss per year over the last 40 years. The data for boreal songbirds are a little less strong because the breeding bird surveys used to generate population trends tend to peter out as you go further away from human population centres. However, many boreal songbirds are very long-distance migrants (i.e. they go all the way from the temperate boreal to South America), which is a big risk factor, and the trends that we do have don’t look very promising for most, including the Connecticut Warbler. Overall, the Connecticut Warbler is declining by about 1% per year. Most of the Connecticut Warbler breeding range is in Canada, which means that Canadians in particular have a responsibility to protect this bird.


This map shows the Breeding Bird Survey trends over the last 40 years, with areas in red declining the most and areas in blue increasing. Overall the Connecticut Warbler gets a ‘medium’ confidence in terms of how real these trends are – this is because there aren’t as many survey routes in their range as we would like, and their density is generally pretty low.

One of the potential threats listed for this species is habitat loss in the winter range. But critically, we have nearly zero information on migration, winter range, winter habitat, or winter ecology. There are only 5 ebird records for this species in the winter range. One in Bolivia, the rest in Colombia. Records in the literature are equally sparse. It’s hard to believe that a North American breeding bird could be so under-studied. Dr. Jay Pitocchelli and colleagues recently (2012) updated the Cornell Lab of Ornithology’s Birds of North America Species Account for Connecticut Warbler, and it is clear there are still huge gaps in our knowledge of this species. For one, it took 70 years after the original specimen was collected for a nest to be described. What we know about breeding biology is largely from a very detailed study of ONE nest in Michigan in the 1960s. Reading through the BNA Account had me more and more intrigued. Then I read this: “Some evidence for nonstop trans-oceanic migration (Monroe 1968, Wetmore et al. 1984) […] Records for West Indies are scarce, but large numbers were reported on Bermuda during Hurricane Emily on 26 Sep 1987 when 75 were grounded; previous daily high had been 3 (Amos 1991).”

Could Connecticut Warblers be trans-oceanic fall migrants, like Blackpoll Warblers? 

Examining the ebird records for the fall migration period were also suggestive of a trans-oceanic flight. Connecticut Warblers are rarely detected along the gulf coast or in Florida, and there is only one ebird record for all of Central America (Panama). In contrast, they have been spotted throughout the Lesser Antilles and into northern South America. Could they be flying over the ocean from the east coast of the US directly to the Lesser Antilles or even direct to South America?


Examining records for Connecticut Warblers shows the potential for a trans-Atlantic flight in fall.

Given the total lack of information on fall migration routes and winter sites, and the potentially mind-blowing fall migration this species might undertake, and the fact that they are a designated stewardship species of the northern forest, they are a prime candidate for migration tracking!

After convincing my supervisor Dr. Oliver Love and mentor Dr. Christian Artuso of Bird Studies Canada that this was worth investigating, last summer I captured over 30 males, and put out tiny 0.5 g data-logging backpacks on 29 of them. These tags are passively recording light levels and hopefully the birds return (with their backpacks intact!) this spring to their breeding sites near Winnipeg, Manitoba, where I will be waiting for them! I suspect they have important stories to tell.

Stay tuned for Part 2: Connecticut Warbler Fieldwork (or, how everything I thought I knew about them was wrong)



Leave a comment

Does what happens in the Tropics stay in the Tropics?

Does what happens in the Tropics stay in the Tropics? Temperate-dwelling people who ‘migrate’ to the Tropics often show ‘carry-over effects’ of their overwintering period upon their return. Those who spent a week or two at an all-inclusive resort might come home relaxed, tanned, perhaps fatter or fitter, depending on the individual preference. I often returned from my tropical sojourns bug-bitten, tired, but elated at the experience. But what about the birds? How does their Tropical stay affect them, once they leave?

We migrate home from Belize on a plane (some times we even get to co-pilot the puddle-jumpers to Belize City!)... Wood Thrush have to wing it there on their own power.

We migrate home from Belize on a plane (sometimes I even get to co-pilot the puddle-jumpers to Belize City!)… Wood Thrush have to wing it there on their own power.

I’ve already blogged about how Wood Thrushes show a decline in body condition over the winter period, ending up in rough shape (at least in Belize) right before they have to leave on spring migration. So what happens on spring migration? Does it matter if they are in poor condition, or in poor-quality habitat in winter? Or do their migration genes just get them on their way, regardless of what condition they are in?

Come back next year, Wood Thrush (with your backpack still on, please)!

Come back next year, Wood Thrush (with your backpack still on, please)!

I tried to answer these questions by combining our geolocator-tracking data (from Belize and also from across the breeding and wintering range) with body condition data and remote sensing of habitat dryness. For the latter, there is a nifty satellite-derived index of habitat quality called NDVI (Normalized Difference Vegetation Index – see figure below). It basically is a picture of the world from space, and the greenness of each pixel in the picture has a value. The more green, the higher the NDVI value, the more productive (as in plant productivitiy), wet, and full of bugs a particular forest is; conversely, less green = less productive, drier, and correspondingly fewer bugs. We know that Wood Thrush body condition is correlated with habitat dryness and bugginess, so this makes the NDVI a good remote indicator of habitat quality.


This is what the Normalized Vegetation Difference Index looks like for the world in March – lots of snow in the North, but green in the Tropics and in the temperate south.

We also have our detailed migration data from geolocators. Just a reminder: these are archival tags (they don’t transmit data) that give us estimates of latitudes and longitudes for each day based on light levels (sunrise/set times and day length). From these data we can determine latitude and longitude, and thus migration timing, i.e. departure date, date crossing the Gulf of Mexico (about the halfway point for Wood Thrushes) and date of arrival at breeding sites. We can also figure out migration distance, speed (overall distance/duration in days), and the number of days spent at stopover sites.

We expected that birds in poor body condition (smaller than expected based on their size) would show negative carry-over effects on migration. For example, a skinny bird might have to delay its departure, or travel more slowly, stopping more frequently along the way, resulting in a later arrival at breeding sites. The breeding arrival date is really critical because this predicts reproductive success in many species. Late arrivers either get crappy territories (males) or crappy mates (females) and which results in fewer babies for those birds in the long run. Of course really poor condition probably results in not surviving spring migration at all! But we can’t measure this because we only have information from birds that survived to bring us back their migration-tracking geolocator. Definitely a bias, but an unavoidable one at the moment.

So we compared body condition of my Belize birds to their spring migration performance. Surprisingly, we found that birds in poor condition didn’t do anything significantly different from those in the best condition! Their timing, stopover behaviour, and speeds were not significantly related to their condition. If you squint hard at the data, it’s almost the opposite effect – birds in good condition hung on later than those in the worst condition!

Wood Thrushes in good shape (body condition positive) were not earlier to leave than birds in poor shape (negative condition index). There was also no difference in timing later during migration, in speed, duration, or distance travelled.

Wood Thrushes in good shape (body condition positive) were not earlier to leave than birds in poor shape (negative condition index). There was also no difference in timing later during migration, in speed, duration, or distance travelled.

Given that this was a pretty small sample size (under 30 birds) all from the same location (Belize), I thought I would expand my tests using some of our other lab tracking data. This is where the NDVI comes in. I had migration data from Wood Thrushes tagged at a breeding site in Pennsylvania, USA – we knew where these guys were in the winter, and using NDVI we could remotely measure the quality of the habitat. I compared  winter habitat quality to spring migration performance to see if birds wintering in drier sites had poor performance compared to birds in wetter sites.

Birds breeding in Pennsylvania wintered in the central part of the winter range, and I looked at the NDVI for each specific winter site to see if drier forests affected their migrations in spring.

Birds breeding in Pennsylvania wintered in the central part of the winter range, and I looked at the NDVI for each specific winter site to see if drier forests affected their migrations in spring.

I found some support for the idea that winter habitat carries-over to affect spring migration. Birds in drier sites departed significantly later than birds in wetter sites; however, the dry-site birds caught up! There was no difference in timing by the midpoint of migration (at the Gulf of Mexico) or by the time they arrived at breeding sites. If these dry-winter-site birds are catching up, they must be moving faster, and sure enough, I found that lower NDVI was associated with faster speeds and shorter spring migration duration. Overall this shows that birds in very dry winter sites may take longer to depart, but they seem to be able to compensate for this en route and arrive at their breeding site without delay.

The last part of my study involved comparing migration performance of birds from my Belize study site, and a site in Costa Rica (La Selva, where co-author Calandra Stanley did the ground-work; she’s now doing a PhD at the Smithsonian/University of Maryland). These two sites are interesting because they are really different in terms of habitat moisture – Belize is always drier, and it dries out faster than Costa Rica. How do these broad-scale differences in habitat quality relate to migration?

Callie Stanley with a Wood Thrush. Her part of this project was the La Selva Costa Rica work, although now she works in Belize on Wood Thrushes.

Callie Stanley with a Wood Thrush. Her part of this project was the La Selva Costa Rica work, although now she works in Belize on Wood Thrushes.

For starters, the Costa Rica birds migrate about 1000km farther on average than the Belize birds! They also depart later and have later timing along their entire migration. Interestingly, the Costa Rican birds go further but stop for about the same number of nights as the Belize birds, indicating that they are more efficient in their migration somehow. This might be because they have significantly longer wings. Overall this goes a long way to explain the species-wide leap-frog migration pattern in Wood Thrushes. We showed in our previous work that Wood Thrushes from the southern parts of the wintering range (e.g. Costa Rica), migrate to the farthest northeast of the breeding range (e.g. up-state New York), essentially ‘leap-frogging‘ over the birds in the middle (e.g. the Belize birds tend to breed in the southeast). The relationship I show with migration and habitat quality could be what’s maintaining this overall, species-level pattern. Birds in Costa Rica can afford to stay longer because the forest is still relatively productive, and their farther breeding sites would still be under snow cover anyway. In contrast, the Belize birds face an increasingly hostile (i.e. dry) winter site. It likely pays for them to get outta dodge earlier and head for their breeding sites in the southeast US, where spring has already sprung. They don’t need to worry about fuelling up as much as the Costa Rica birds either, because they are covering a lot less ground.

Wood Thrushes in Costa Rica are in wetter habitat, and they migrate on average 1000km further than Wood Thrush from drier habitat in Belize.

Wood Thrushes in Costa Rica are in wetter habitat, and they migrate on average 1000km further than Wood Thrush from drier habitat in Belize.

So overall, what happens in the Tropics doesn’t seem to have much of an effect on individual Wood Thrushes (and where it does, they seem to be able to compensate by speeding up their migration). But the overall patterns of habitat quality (drier in the north, wetter in the south) are correlated with broad-scale differences in migration behaviour (earlier timing, shorter distances in the north; later timing, longer distances in the south) leading to the leap-frog system that we see at the species-level.

If you think this is as interesting as I do (although I admit I’m biased!), you might want to read the whole paper, found here:

McKinnon, EA, Stanley, CQ, and BJM Stutchbury. 2015. Carry-over effects of nonbreeding habitat on start-to-finish spring migration performance of a songbird. PLOS ONE.

Leave a comment

Why young birds don’t get the worm

Songbird migration is amazing.

Imagine this:

You hatch somewhere in eastern North America, let’s say… a nice sugar maple forest in Vermont. You hang out in your nest for a while, happily eating whatever your parents shove down your throat, jockeying with your siblings for space and food, until one day – you hop out with a klutzy fluttering of wings. You then follow Mum or Dad around for awhile, maybe joined by one or two of your four sibs. Eventually they wander off, and you are on your own, happily gorging on berries, growing in your last feathers. Finally, the days get shorter and something in your brain clicks and you know it’s time to move. Without any guidance, you simply take off one evening and point your beak southwards. Days or weeks later (and thousands of kilometres), you arrive in a steamy tropical jungle that somehow, feels like home. 

As spring arrives (probably detected by subtle changes in day length) you get that familiar urge to fly, only this time, it’s northwards. You have no idea of your exact destination, just a vague sense of where you were hatched (must be a good place to breed around there, you survived, after all, right?). You head northwards only to be stopped in a few days by an immense body of water, with no land in sight on the other side. This is the Gulf of Mexico. From the tip of the Yucatan penninsula of Mexico, it’s nearly 1000 km to the U.S. coast on the north side. Yet instinct tells you to go for it – so one evening, you launch straight out over the open water. Hours of flying later (likely well into the next day), you finally spy a wavering outline of land ahead. Soon you touch down on a windswept barrier island at the mouth of the Mississippi, near New Orleans. And this is just the start of your first spring migration.

Wood Thrush - an amazing migratory songbird!

Wood Thrush – an amazing migratory songbird!

I’ve tried to walk you through the first migrations of a songbird like the Wood Thrush, because I think it’s so hard for us thinking apes to get our heads around such insane-seeming instinctive behaviours. Songbirds are really like little programmed robots – their genes are so finely tuned that they can accomplish these amazing feats of migration without ever thinking about it. For first-time migrants, this is even more amazing, since they have no opportunity to learn their routes from their parents, or other adults. In fact, since most songbirds migrate at night, they probably can’t even see the other birds they might be flying with.

Scientists have studied the development of migratory behaviour in the lab, and found that while a lot of it is pure instinct, there is an important contribution of experience. Nestlings raised in captivity start getting the migration fidgets (the academic term for this increased hopping and fluttering around in their cages is zugunruhe, German for ‘nocturnal restlessness’) at the right time to start their migrations southwards. They also know what general direction they should go, i.e. southwest. Lab studies have shown this by the simple but ingenious ‘Emlen’ funnel – a paper funnel with an ink pad at the bottom. Put a bird in (and some screen over the top) and the bird will hop all night in the direction it wants to fly, each time stamping its feet on the paper funnel, which thus records the direction the bird wants to go. But young songbirds do have trouble if they get blown off course. Adult birds figure it out and re-orient the following night, but juveniles keep doggedly on the same course. This is why fall is a good time to see rare birds – juveniles are moving around and sometimes end up in places they shouldn’t be. Presumably, natural selection takes care of any juvenile that gets too far off track with the result that juveniles on their first migration probably have pretty high mortality rates.

By the time birds undergo spring migration, all the juveniles have by definition survived fall migration and spent the winter in an appropriate place. In spring, juvenile birds can re-orient themselves when they are blown off course, so the experience of migrating southwards in fall must have given them some sort of overall map in their brains. However, the exact route that they need to take is not the same as in fall (many birds do a loop migration, where spring and fall migration occur along different routes, probably because of favourable wind patterns). This means juvenile birds have to figure out a new route to get to their inaugural breeding site. Songbirds tend to be site faithful to the same territories after they have bred there once, but juveniles rarely return to the exact territory where they were hatched (that could lead to inbreeding). Instead, juveniles are thought to aim for the general area where they were hatched (leading to the patterns of migratory connectivity we discovered), so that a young bird from Vermont might return somewhere nearby – a few hundred kilometres away. It probably wouldn’t breed in Indiana, for example, but might end up in New York state. So even in spring, when juveniles have a little experience, it’s still pretty amazing that they can find their way back to a breeding site at all. To make things even more challenging, if you are a Wood Thrush, as in my example above, in spring you most definitely want to take the most efficient route back to the breeding grounds, which means dealing with the 1,000-km open-water crossing of the Gulf of Mexico.

I studied the spring migrations of juvenile Wood Thrushes from my study site in Belize ( and also used some data collected by my colleague Callie Stanley during her Masters work in Costa Rica (at La Selva Biological Station). One of the many advantages of studying birds in Belize is that I could catch juvenile birds before they left on migration, and if they survived to return the following year, I could map their very first journey north. To do this I used little bird backpacks called geolocators to track their migrations. See my previous blog post for an explanation of how they work. Basically, they record where the bird is each day and I have to recapture the same individual one year later to get the data.

Come back next year, Wood Thrush (with your backpack still on, please)!

Come back next year, Wood Thrush (with your backpack still on, please)!

It’s challenging enough to catch those ‘golden’ backpack-wearing birds, but the odds of getting the juveniles (now returning as adults) is even lower. Most juveniles just don’t make it to breeding sites and back. Where exactly most of them get into trouble, we don’t know. Could be they choose their tail winds poorly and get stuck out over the Gulf of Mexico. Maybe they don’t have a healthy fear of cell towers or glass skyscrapers and meet an untimely end that way. Until we have backpacks that transmit the data remotely, we won’t know what happens to all the birds that don’t come back.

After several years, I ended up with a pretty decent sample of 17 first-time spring migration tracks for Wood Thrushes. It’s not a lot, but this is the first time songbirds of any species have been followed from start-to-finish on spring migration! So what do they do?

First of all, they leave late.

The first-time migrants hung out at their tropical wintering sites for almost a week more than adults! One idea was that maybe they are in rough shape after duking it out for food with adult birds the whole winter. So I looked at the body condition of adult versus juvenile birds at my site in Belize: no difference. In fact, the juveniles were in a bit better condition than adults (not significant though). Scratch that idea! So why are they leaving late?

One clue is that not only did they leave late, they got more and more behind the adults as they headed northwards. By the time they arrived at breeding sites, juvenile Wood Thrushes were almost two weeks behind adults!

This is one bird we tracked twice, once as a juvenile and the next year as an adult. Check out how much earlier he arrives when he's an adult! 26 April versus 11 May as a juvenile.

This is one bird we tracked twice, once as a juvenile and the next year as an adult. Check out how much earlier he arrives when he’s an adult! 26 April versus 11 May as a juvenile.

The juveniles start to get more and more behind because they stop more frequently in the U.S. as they travel northwards. Why would they do this? Maybe they have to, if they are in rough shape (although I suspect not). It could be that they are less efficient at flying (they do have shorter wings) or that they have less experience selecting tail winds, so each flight doesn’t take them as far as adults. However, I also found that juveniles were just as likely as adults to cross the Gulf of Mexico, and they didn’t stop for longer before (to prepare) or after (to recover), which seems to suggest that they can perform as well as adults.

One idea is that juvenile birds might actually be programmed to arrive later. There are big costs to arriving at a breeding site first: it could get cold, food could be limiting, and early birds will likely have to fend off more than one rival for that prime territory. In contrast, birds that wait a bit arrive when all the adults have settled on the best territories, and there may be comparatively little fuss when they arrive and set up shop in a lesser territory nearby. The benefits might be that the weather is better and therefore food is probably more predictable, and a later arriving bird might not face as many risky territorial challenges. Later arriving birds may not get the best territory (or mate), so their reproductive success might be low, but maybe instead they ‘prioritize’ making it through their first breeding season alive. If you were a juvenile bird, I suspect this later-arrival strategy could be your gene-driven game plan.

We don’t really know, of course, why juvenile Wood Thrushes took a more leisurely spring migration. But now we know how – they both leave late, and stop more on their way northwards!

Read our full paper (Open Access!) here (email or tweet me if you can’t get it):

McKinnon, E.A., Fraser, K.C., Stanley, C.Q., and B. J. M. Stutchbury. 2014. Tracking from the Tropics reveals behaviour of juvenile birds on their first spring migration. PLOS ONE.



Connecting breeding and wintering sites for a declining migratory songbird

We are losing our migratory songbirds. It’s a fact, and there are many intertwined possible mechanisms, including habitat loss, climate change, invasive species, chemical and light pollution, etc. But for conservation practitioners trying to save the songbirds, there is a gaping hole in our understanding of their biology. We do not know where most small songbirds go when they leave their breeding sites.

Take the Wood Thrush, for example. It breeds in eastern North America, where it has been studied for decades. We know that forest fragmentation and acid rain definitely have effects on breeding Wood Thrushes: birds produce fewer young in small forest patches, or where acid rain has depleted the calcium from the soil and therefore lowered the amount of insect food. But every fall, Wood Thrushes take off from their breeding sites and head to southern Mexico and Central America. Each Wood Thrush heads for a patch of tropical forest somewhere between Veracruz, MX, and the Panama canal. That’s an area of over 500,000 square kilometres! This is why bird banding doesn’t work for making connections for most small birds (unless they are really rare and range restricted, e.g. Bicknell’s Thrush). Finding a Wood Thrush wearing a leg band (marking its breeding site) is like a finding a needle in a haystack, only the haystack covers thousands of square kilometres and it’s full of other needles with no leg bands!

Why is it so important to know where each breeding population goes? For starters, there are geographic patterns in the population declines. Wood Thrushes in Canada (and in the north-east of the U.S.) are disappearing faster than they are in the central and western parts of their breeding range. Are these different breeding populations experiencing different threats on migration or at their winter sites? Do they all mingle on the winter grounds, i.e. a bird from Ontario hangs out with breeders from Georgia? Or do different populations have distinct wintering ranges, i.e. all the Ontario birds hang out together with other Ontario birds? We call this idea of sharing neighbours ‘migratory connectivity’. If it’s strong, the birds stick with their breeding buddies in the winter. If migratory connectivity is weak, the birds might be found next to any ol’ Wood Thrush regardless of breeding origin.

Wood Thrush breeding trends

Breeding grounds population trends for Wood Thrushes measured by Breeding Bird Surveys from 1966-2012. The major breeding regions we used in our study are shown by the dashed lines – Northeast (NE), central east (CE), Southeast (SE) and Midwest (MW).

The most effective way to figure out patterns of migratory connectivity out is to follow these birds on migration.

Wood Thrushes easily fit into the palm of your hand, and they weigh less than a tennis ball. How do you follow one over 4,000 km of migration? Researchers have been trying for years to indirectly track birds using chemical markers in their tissues, or DNA structure, or even just by banding lots of individuals in one location and hoping that a bird is captured somewhere else. For Wood Thrushes, not one of these techniques has worked. Despite thousands of Wood Thrushes banded, only one has ever been recaptured in the opposite season. This bird was banded by my colleagues in Nicaragua and hit a window of someone’s house in Pennsylvania in 2011 (read full story here). Thankfully someone noticed the thump on the window and the leg band that identified this bird. But one band recovery out of thousands is not enough to paint a full picture of migratory connectivity for Wood Thrushes. Chemical markers have been somewhat more successful in making connections for Wood Thrushes at a very broad scale (read about it here). But it really was the miniaturization of tracking devices called ‘geolocators’ that revolutionized our understanding of Wood Thrush migratory connectivity.

Light-level Geolocator

Geolocator harnessed for a Wood Thrush. The stalk is at the back and points to the bird’s tail; the white square at the end is the light sensor. The super-strong but soft teflon straps are adjusted to fit each bird.

Wood Thrush wearing a geolocator. Only the tip of the light stalk pokes through the feathers once the geolocator has settled on the bird.

Wood Thrush wearing a geolocator. Only the tip of the light stalk pokes through the feathers once the geolocator has settled on the bird.

Geolocators are tiny devices (<2 g) that can be attached to a bird like a backpack, except they go over the legs, and not the wings. These devices are very simple: battery + clock + light sensor + chip to record the data. Before you put the tag on the bird, you turn it on and program it with the current time. Once on the bird, it records light levels continually relative to the clock. If the bird moves east, sunrise will be slightly earlier. If the bird moves south in the fall, day length will be longer. By inputing the recorded times of sunrises and sunsets into a computer program, you can estimate the latitude and longitude where the bird was each day. Easy, right? Well not quite. The most challenging thing about these tiny geolocators is that THEY DO NOT TRANSMIT DATA. This means that we put the backpack on the bird, it migrates thousands of kilometres, does its thing in Mexico or Central America for the winter, migrates back in spring, THEN we have to catch it again to take the backpack off to get the data. Seems nearly impossible, but it does eventually work.

The culmination of years of this type of tracking, and hours and hours of effort by graduate students, field techs, volunteers, and of course our project leader, Dr. Bridget Stutchbury, is a map.


Breeding-wintering connections for Wood Thrushes. Each star is a site where geolocators were deployed on Wood Thrushes, and the round circles are the birds’ sites in the opposite season. Each deployment location is colour coded. Inset photo shows a Wood Thrush with a geolocator.

Not just any map. This map contains detailed migration data from over 100 Wood Thrushes tracked from 7 breeding sites and 4 winter sites. It tells us exactly where each bird goes, and what route it takes to get there. This is the first time a migratory connectivity map has been produced for a songbird using tracking from both breeding and winter sites (although our lab has done some pretty awesome work with Purple Martins too).

What did we discover? First of all, there is a pattern. Birds from Canada don’t usually hang out with birds from Georgia in the winter. They stick with their buddies from the central and north-east, and hang out in Nicaragua and Honduras. In contrast, Wood Thrushes I tracked from Belize all head to the central and south of the breeding range: Kentucky, Virginia, the Carolinas, a few even bred at the very southern limits of their range in Florida. Overall we call this pattern ‘leap-frog’ because the birds breeding the farthest north actually migrate the farthest south, ‘leap-frogging’ over the southern breeding populations. The connections also tended to be predicted by longitude, so that birds breeding further east (and north) spent the winter further east (and south). So I can tell you that if you are Canadian visiting the Mexican riviera on vacation and you see a few Wood Thrushes – odds are these are not fellow Canucks, but probably birds from the southern US. If you want to see your ‘Canadian’ Wood Thrushes, you would have to head further south – the Mosquito Coast of Nicaragua would be a good option (a little more adventuresome for a vacation too!).

We also discovered some amazing patterns in migration. In fall the Wood Thrushes tend to funnel south along three major routes – either through Florida then island-hop over Cuba to Honduras, 2) cross the Gulf to the Yucatan peninsula diagonally through the Florida panhandle, or 3) cross the Gulf to the Yucatan from Louisiana. The choice of route was generally predicted by where the birds were breeding, i.e. eastern breeders took the eastern (Florida) route, while western breeders were more likely to cross the Gulf from Louisiana.


Fall migration routes for Wood Thrushes colour-coded by breeding region. Width of line shows proportion of the entire species using that route.

Fall migration routes for Wood Thrushes colour-coded by breeding region. Width of line shows proportion of the entire species using that route.

In spring, it’s a much more interesting story. We found that almost 75% of ALL Wood Thrushes cross the Gulf of Mexico from the tip of the Yucatan peninsula to land in a small area of Louisiana on the northern gulf coast. That means almost the entire global population of this bird uses that one tiny piece of land near New Orleans every spring!

Spring migration routes for Wood Thrushes colour-coded by breeding region. Width of lines shows proportion of the entire species that uses that route.

Spring migration routes for Wood Thrushes colour-coded by breeding region. Width of lines shows proportion of the entire species that uses that route.

This is why our work is so important for conservation. We know now where the ‘hotspots’ are that are used by the most Wood Thrushes at a global scale, and we also know which areas are important for specific breeding populations. For example, if you want to protect habitat for those Canadian Wood Thrushes – invest in shade-coffee and sustainable forest use programs in Nicaragua. Better yet, contact the local ministry of the environment (MARENA) and figure out how you can help conserve forests in Nicaragua. The truth is, they aren’t really Canadian Wood Thrushes after all – they have duel citizenship!

Our full paper is published in Conservation Biology. If you can’t access it, email me: emilymckinnon12 AT or contact me on Twitter @BirdBiologist and I’ll send you a pdf.

Stanley, C. Q., E. A. McKinnon, K. C. Fraser, M. P. MacPherson, G. Casbourn, L. Friesen, P. P. Marra, C. E. Studds, T. B. Ryder, N. Diggs, and B. J. Stutchbury. 2014. Connectivity of Wood Thrush breeding, wintering, and migration sites based on range-wide tracking. Conservation Biology Early online.