The winter of the Snowy Owl

Back in late November, eBird reviewer Sam Galick got us ready for the Snowy Owl invasion of 2011-2012. For three months now eBirders across the US and Canada have been reveling in these striking tundra emissaries, which have appeared from coast to coast, many well south of their normal range. eBirders Jesse Ellis and Skye Haas have mined the email reports and listservs to ensure that no Snowy Owl report has escaped eBird’s net. Thanks to their efforts, and submissions from thousands of eBirders worldwide, this surely has been the best documented Snowy Owl invasion in the history of ornithology. With three months of data under our belts, and 2-3 months left to enjoy these majestic raptors before they head north, this is a great time to compare this winter’s irruption to years past. By any metric, this is one of the larger invasions, but how big is it? What do the geographical patterns tell us? Below we explore this winter’s Snowy Owl invasion within a historical context. But first, here’s a quick quiz: how many US states have never recorded a Snowy Owl? Which ones?

As an introduction, for those that have not yet had the good fortune of having Snowy Owls nearby, check out the Birds of North America Online species account and Gerrit Vyn’s surreal photo and video essay. His magical videography shows the owls perching on the windswept driftwood at Ocean Shores, Washington, and his narration captures both the mystique and the basic biology of these amazing creatures. In the video, multiple birds are often sitting just meters apart, and indeed, this area has been one of the hotspots for Snowies this winter, with perhaps the largest concentrations (up to 13) for a single spot in the Lower 48. Just to the north, the Vancouver area has had even more (up to 31). How unusual is this winter? Below we summarize the emerging patterns region-by-region.

2011-2012 Snowy Owls

Check out eBird’s Snowy Owl map — November 2011 to February 2012: Toggle the “Show Points Sooner” box on the right to see more precise points of occurrence and click to see specific checklists. Then compare it to the November map to see how it has filled in!

Continent-wide, it is clear that a massive movement is afoot, with Snowy Owl sightings in 31 states and almost all Canadian provinces. A few patterns quickly become obvious. When you click “show points” on the above map it is especially striking how Snowy Owls occur on the beaches, valleys, and plains, but totally eschew the mountains. This is nothing new–Snowy Owls are birds of flat, open country–but to see it shown so precisely on a map is definitely striking.

Notice the geographic patterns in places where we have data. Viewing the map in the grid mode helps you distinguish between areas with data but no owls (the gray grid cells) and the areas with no data (the blank areas, without gray or purple shading). This is important, and it highlights the need for more information from parts of Nevada and northeast Arizona, along with Wyoming, eastern Montana, and the Dakotas. Some of these latter areas surely have Snowy Owls, but lack eBirders.

Notice the geographic patterns in places where the owls are being found. As noted by Sam Galick, most birds in the eastern US have been occurring along shorelines (e.g., Great Lakes, the Saint Lawrence River, and Atlantic coast), whereas birds in the upper Midwest have been more evenly spread across the landscape. Although the Upper Great Plains is our sparsest data region, it is clear there are lots of Snowy Owls there. The darker purple cells are the ones where Snowy Owls are being found on a higher percentage of checklists. If you look closely at the map, you’ll be able to see five shades of purple–the full gamut–meaning that the darkest cells (in eastern South Dakota, for example) have Snowy Owls reports on >40% of the checklists. Think about that for a minute! Some of the checklists in that area are probably from forested areas, and urban backyards, and other areas that lack Snowies. Within appropriate habitat, it could very well be hard to miss a Snowy Owl in South Dakota right now! Maybe that’s an exaggeration, but this 42 mile traveling count reported 20 Snowies. Wow!

Other areas of concentration are on the Pacific Northwest coast (south to central Oregon), eastern Montana, the Prairie Provinces of Canada, and the Great Lakes. The most southerly Snowies have been in the center of the country, where they have reached northern Oklahoma and central Arkansas. On the East Coast they have not shown up south of New Jersey and Pennsylvania, but there have been many north of there.

Comparing the past five years

On a continental scale, how can we compare this year to past years? Remember that eBird line graphs give a lot of additional information that can’t be seen in the maps. If we make a line graph for all of North America, we can see how the overall continent-wide frequency this year compares to the past five years (just check off the “Separate years (max. 5 years, 1 species)” box after checking the “change date” option. This displays a graph for Snowy Owl occurrence from October to December for the years 2007-2011. Here, it is obvious that 2011 has a frequency of up to 1.8%, well above past years (which are less than 0.1%), although fall 2008 climbed to nearly 1%. Click on here to explore this eBird output, alter the dates, or explore other metrics such as abundance or high counts. To explore output for January-March in the years 2008-2012, click here; notice how the mid-winter time frame of 2009 is actually surpassing 2012 in both frequency and abundance!

This last fact may be surprising, and is worth delving into a bit deeper. Below we discuss these regional patterns in more depth, and draw on the historical ornithological literature to put the records in broader context.

Best winter ever, or not? How big is big?

There have been some casual claims that this is “the best winter ever” for Snowy Owls. But a short walk through the historical literature, and a bit of perspective, would temper those claims. Without question, this winter is one of the biggest invasions. And with eBird, it is surely the best documented invasion in ornithological history, and it may even end up with the most Snowy Owl sightings (many pertaining to the same bird seen by many birders). This winter has not only benefited from the Information Age (i.e., eBird!), but also from an increase in the number of birders and naturalists around to notice the birds. For these reasons, comparing metrics like the number of Snowy Owl sightings this year vs. past years is dangerous (but we’ll launch into it anyway since it is one of the easier metrics to look at)! However, it is probably safe to say that when aggregated reports from winters past match or exceed this year’s totals, then by today’s standards those past years must have been even bigger than the historical calculus would indicate.

In the Northeast we can immediately say that this winter is an irruption, but not a very big one. In the Northeast, Snowy Owls invade once every 3-6 years or so, and many invasions have involved many more birds than are being found this year. A glance at the map for the Winter 2008-2009 shows that even based on past eBird records (which were about 44% of today’s volume), there were more Snowies around in 2008-2009. Looking farther back, in 1926-27 there were 294 birds in Massachusetts alone (Veit and Petersen 1993, Birds of Massachusetts); numbers simply unheard of in recent winters. A few times in the last 40 years double-digit counts have occurred at single locations on the Massachusetts coast, but not this year. This year’s total of 21 banded at Logan Airport (reported by the New York Times), most of them moving through, is still well below the 43 banded in 1986-87. Even with sparse eBird reporting a decade ago, look how far south they got in 2001-2002.

In the Pacific Northwest it’s the same story. The really big years in Washington have recorded hundreds of owls (>1000 in 1916-1917 and >200 in 1973-1974 according to Wahl, Tweit, and Mlodinow’s 2005 Birds of Washington), and the best winters in Oregon have seen double-digit counts in a single CBC circle, well above this year’s numbers. When Snowy Owls reach California, it is a sign of a really big winter on the West Coast, and there were at least 10 there in 1916-1917, and 32 (!) in the historic winter of 1973-74. No Snowies have reached California yet this year (although singles did in Jan 2006 and Dec 2006-Jan 2007).

Elsewhere at the fringes of the species’ range, this year’s invasion has yet to make much news. The one bird found in Hawaii did provide an amazing new record for that state. A bird in Arkansas and another in Oklahoma have been noteworthy, but in the East there have not been any birds remarkably far south.

Why then is this year such a big deal? It is because the number of owls being found in the Midwest and Great Plains this winter has been truly epic! The numbers from Kansas and Missouri (thanks to Mark Robbins for compiling these) have been most revealing. The best previous year in Kansas was 81 birds in 1974-1975, but this year there have been 101 (and counting)! In Missouri, the previous single-year high was 8, but this year has seen 54 (and counting)! The 1974-75 winter may well have been better in Kansas, but the sevenfold increase in records in Missouri may be more than could be explained by more observers and better information sharing. Mark reports that the birds in Kansas have been concentrated in the eastern part of the state (also visible on the eBird maps), so the southern wedge of their push has been near the Kansas-Missouri state line. Nebraska, Iowa, Montana, the Dakotas, Minnesota, Michigan, and Wisconsin seem to define the core area of occurrence this winter, and by all accounts the numbers here are off the charts.

Are they the best ever? Hard to say without some measure of effort expended collecting the observations (as in the above eBird line graphs), but there is no question that this winter is unusual and one of the best ever. When the Christmas Bird Count data are all in and available, that will help tell the story, since those data do include measures of effort surveying for Snowy Owls (and all birds) that go back many decades.

Age, sex, and mortality

Although it is generally true that older owls and male owls tend to be whiter, and younger birds and females tend to be darker, assigning any individual bird to age and sex is very hard to do, and probably impossible. There is enough individual variation that some old females can be very dark and others almost pure white, so even broad generalizations may not be safe. This makes it hard to know to what extent the high numbers this winter are from unusually high numbers of immatures vs. unusual dispersal of all age classes. Disparate data points on banded birds and birds found dead may help with such questions.

In Kansas and Missouri, Mark Robbins is performing necropsies on the dead birds he receives, and can report that most have been immature birds that have been food-stressed (or starving), but that so far cause of death has always been collisions (with cars, powerlines, and even trains!). There is no question that birds turning up this far south have a difficult time finding enough food to survive the winter and make the long trip back north.

Final thoughts

These Snowy Owl invasions have been recognized for over a century, and some of the older anecdotal reports of thousands of Snowies in a single state make one wonder about how much our environment has changed in the interim. This year is impressive, but would a comparable invasion 100 years ago have had even more owls? Is climate change to blame for apparently reduced numbers during these “big” invasions? Are the same number of owls around, but perhaps wintering farther north than they were able to 100 years ago? Many of these questions are unanswerable, given how hard it is to survey Arctic birds in their harsh environment. But they are fun to think about nonetheless.

The high ratio of immatures probably means that high productivity in the Arctic was an important driver of this year’s invasion. Given the areas of concentration, the source for these birds was presumably the central Arctic, or central-western Arctic. But without data from the breeding grounds, this connection is hard to make and even harder to prove.

We’ll be sure to post a year-end wrap-up of this invasion, building on this story and getting a little deeper into patterns of the past and how this year compares, as well as some of the theories as to why these invasions occur. In the meantime, let’s keep reporting the owls we see. Often owls continue to push south during February, so perhaps the two states without Snowy Owl records (Arizona and New Mexico, to answer the quiz in the introduction) will finally get a record. Snowy Owls have reached Bermuda in the past–could one ever make it to the Bahamas? Mexico?

For the good of the owls

And in closing, please re-read the excellent cautions by Sam Galick against disturbing or stressing these beautiful owls. They have traveled a great distance and many of the more accessible and approachable birds are also likely stressed for food. Please keep a respectful distance.

More on Snowy Owls

Media: Snowy Owls have been a media sensation as well, covered by the Syracuse Post-Standard, the Bangor Daily News (also here) and even the New York Times (among others). Our favorite (again) is this photo and video essay filmed and narrated by Gerrit Vyn of the Cornell Lab.

Scientific Literature:

Shelford (1945) discusses the possible relationship between lemming cycles and Snowy Owl invasions.

Kerlinger et al. (1985) give some alternate ideas about what causes these irruptions (Kerlinger, P., M. R. Lein, and B. J. Sevick. 1985. Distribution and population fluctuations of wintering Snowy Owls (Nyctea scandiaca) in North America. Can. J. Zool. 63: 1829-1834)

Snowy Owl on West Dennis Beach, MA by Jeremiah R. Trimble

Snowy Owl at Dennis, Massachusetts. Photo by Jeremiah Trimble.