Science Behind The Dabbling Duck Migration Forecast

Mid-latitude Duck Migrations (Missouri, Indiana, Memphis)

Because of concerns about mallards not showing up in the same numbers or as early as they once did at southern latitudes of North America, researchers sought to better understand weather factors causing migration of these ducks during autumn and winter.

In 2010, Schummer and colleagues published an article in the Journal of Wildlife Management (vol 94, pages 94-101) where they described an equation, called a Weather Severity Index, or WSI, that explained significant variation in migration of dabbling ducks through Missouri (and presumably eastern North America).

Although hunters and others interested in duck migration had long known that cold temperatures and snow cover eventually caused migration by mallards and other dabbling ducks, Schummer and colleagues were the first to develop a WSI that enabled calculation of a value to more accurately determine specific timing of migrations by dabbling ducks.

Cumulative WSI = – [average daily temperature oC ] + number of consecutive days with average temperature <0 o C + snow depth + number of consecutive days with snow cover

NOTE:  first change the sign of average daily temperature in oC before adding it to other Cumulative WSI variables, this ensures that accumulated Cumulative WSI values increase with increasing weather severity.

As Cumulative WSI values approach 8 for mallards and zero for other dabbling ducks, there was increased likelihood of migrations by these ducks from the latitude of Missouri to southern locations. At values above Cumulative WSI thresholds, rate of change in relative abundance of ducks became increasingly negative indicating increased likelihood of migration to southern latitudes.

Northern Latitude Migrations (Great Lakes – Prairie and north)

Continued interest in determining autumn-winter weather events that cause waterfowl to migrate and with funding from the Gulf Coastal Plains and Ozarks Landscape Conservation Cooperative, we continued the WSI research at broader geographic scales.  Lena Vanden Elsen, an MSc candidate at the University of Western Ontario, and Dr. Schummer, Long Point Waterfowl Scientist, are expanding the WSI research to investigate factors influencing migration out of northern staging areas (e.g., Iowa, Long Point) into mid-latitude locations (e.g., Missouri, Virginia), thereby refining our prior results from the limited geographic range of Missouri.

Recently, Lena conducted analyses similar to Schummer et al. 2010 using waterfowl survey data from Long Point, 1979 – 2005.  Lena’s MSc research will be conducted using waterfowl survey data from throughout the Mississippi and Atlantic Flyways of North America.  Notably, Lena is investigating species-specific WSIs for all dabbling ducks rather than the two categories of mallards and ‘other dabbling ducks’ as presented in Schummer et al. 2010.

Preliminary results from Lena’s analyses of the Long Point data suggest that migration thresholds differ between northern (Long Point) and mid-latitudes (Missouri).  We determined that migration of dabbling ducks from Long Point to southern latitudes was best explained by the 14-day mean temperature (TempMean WSI) and that this threshold was, on average, met nearly 2 weeks before the Cumulative WSI threshold reported in Schummer et al. 2010.

14-day mean temperature (TempMean WSI) thresholds by species are:

Green-winged teal = 11°C

American wigeon = 10°C

Northern pintail = 10°C

Gadwall = 8.5°C

Mallard = 8°C


At 14-day mean temperatures below these thresholds our model predicts that abundance of the respective species will begin to decline at Long Point and presumable increase at more southern latitudes.  In contrast to our analysis from Missouri where it often took below freezing temperatures and snow to cause a mallard migration, we determined that mallard abundance began to decline at Long Point well before snow accumulation.  In fact, few surveys for any species investigated indicated that duck abundance increased when mean temperatures were ≤ 0°C.  Again, these results suggest that dabbling ducks leave northern latitudes such as Long Point at lesser weather severity thresholds than at mid-latitude staging areas.  We hypothesize this occurs because likelihood of long-term freezing and snow are greater at northern latitudes and thus, the “wait-it-out” strategy used by ducks at mid-latitudes may not apply at the Great Lakes and other northern staging areas.


Therefore, as an interim threshold for northern latitudes we will be using the TempMean WSI for 2012-2013 migration forecasts (Vanden Elsen and Schummer, unpublished results).  For our mid-latitude locations (i.e., Missouri, Indiana, and Tennessee) we continue to use the published Cumulative WSI (Schummer et al. 2010).


Please read more about our research at the links below:

Comments are closed.