Billions of animals migrate to track seasonal pulses in resources. Optimally timing migration is a key strategy, yet the ability of animals to compensate for phenological mismatches en route is largely unknown. We studied a population of mule deer (Odocoileus hemionus) in Wyoming that lack reliable cues on their desert winter range, causing them to start migration 70 days ahead to 52 days behind the wave of spring green-up. By adjusting movement speed and stopover use, however, individual deer arrive at the summer range within an average 6-day window. Late migrants move 2.5 times faster and spend 72% less time on stopovers than early migrants, which allows them to catch the green wave. Ungulates, and potentially other migratory species, possess cognitive abilities to recognize where they are in space and time relative to key resources. Such behavioral capacity may allow migratory taxa to maintain foraging benefits amid rapidly changing phenology.

Animal capture and handling

From 2014–2020, we captured n = 220 adult female mule deer (>1-yr-old) in the Red Desert near Rock Springs, Wyoming, USA (41° 35′N, 109° 12′W) as part of a long-term study. We recaptured deer each March and December for a total of n = 528 animal-years of data. All deer were captured via helicopter net-gunning^1,2. Mule deer in this portion of the Sublette Herd migrate a variety of distances to their summer ranges in northwestern Wyoming^3,4. Herein, we focused on n = 72 long-distance migrants (n = 152 animal-years) that migrated 134–293 km and spent the summer north of Pinedale, WY (42° 51′N, 109° 51′W). During captures, we used an electronic platform scale (± 0.1 kg) to measure body mass (kg) and a portable ultrasound (Ibex, E.I. Medical Imaging, Loveland, CO) to measure maximum rump fat (mm). Following previously applied methods⁵, we used body mass, maximum rump fat, and a body-condition score to estimate percent-scaled ingesta-free body fat (IFBFat)^5,6. For captures in March, we used an ultrasound to determine pregnancy, including fetal rate (number of fetuses per deer) and fetal development via measures of the fetal eye diameter (mm). To estimate the age of each deer, we extracted the lower right incisiform canine and used cementum annuli aging technique^7–9, which was conducted by the Matson’s Laboratory in Manhattan, Montana, USA. From 2014–2020, we outfitted all deer with store-on-board or iridium GPS collars that collected locations every 1–2 hrs (Advanced Telemetry Systems, Isanti, MN, USA; Lotek Wireless, Newmarket, ON, CAN; Telonics, Mesa, AZ, USA). We also included GPS collar data from a previous study on the Sublette Mule Deer Herd (2011–2013)⁴ to analyze movement for n = 27 additional deer (n = 66 animal-years), which were outfitted with store-on-board GPS collars that collected locations every 3 hours (Telonics, Mesa, AZ, USA). All animal capture and handling protocols were approved by the Wyoming Game and Fish Department (Chapter 33-937) and an Institutional Animal Care and Use Committee at the University of Wyoming (20131111KM00040, 20151204KM00135, 20170215KM00260, 20200302MK00411).

Delineation of migratory routes and seasonal ranges

We used Net Square Displacements (NSDs¹⁰) to determine the timing of spring and autumn migration, delineate migratory routes, and determine the net displacement (km) between the start of spring migration and each GPS location along the migratory route. We determined winter range use for each deer by extracting GPS locations between the end of autumn migration and the start of spring migration (or between the time of capture and start of spring migration if the end of autumn migration was unknown). We determined summer range use for each deer by extracting GPS locations between the end of spring migration and the start of autumn migration (or between the end of spring migration and time of collar failure or mortality on summer range). We used a 95% Kernel Utilization Distribution (KUD¹¹) to delineate the winter range (41.63 ± 7.26 km² [x̄ ± 95% CI]) and summer range (7.26 ± 1.61 km²) of each animal-year. We removed 0.10%, 0.09%, and 0.24% of all GPS locations during migration, on winter range, and on summer range, respectively, because the movement rate between consecutive locations was greater than 10.8 km/hr and indicated an inaccurate GPS fix.

Green wave surfing

We evaluated the ability of mule deer to track green-up of plants during spring migration by analyzing the synchronicity between movement and peak IRG. We determined IRG by extracting the first derivative of double-logistic curves that were fitted to the annual time series of NDVI¹². Days from peak IRG (hereafter referred to as Days-From-Peak) were calculated as the difference in days between the date of every GPS location for a deer and the date of peak IRG at the same GPS location^12,13. A theoretically perfect surfer occupies a location on the same day that peak IRG occurs (Days-From-Peak = 0)¹². We calculated mean Days-From-Peak and IRG for each day and kilometer of an individual’s migration to reduce pseudoreplication and account for irregular GPS fixes^12,13. We quantified an individual’s location on the green wave at the start and end of spring migration by calculating the difference in days between the mean date of peak IRG on each seasonal range and the date an individual departed their winter range or arrived at their summer range.

References

1. Barrett, M. W., Nolan, J. W. & Roy, L. D. Evaluation of a hand-held net-gun to capture large mammals. Wildlife Society Bulletin 10, 108–114 (1982).
2. Krausman, P. R., Hervert J. J. & Ordway, L. L. Capturing deer and mountain sheep with a net-gun. Wildlife Society Bulletin 13, 71–73 (1985).
3. Kauffman, M. J. et al. Wild Migrations: Atlas of Wyoming’s Ungulates (Oregon State University Press, Corvallis, OR, 2018).
4. Sawyer, H., Middleton, A. D., Hayes, M. M., Kauffman, M. J. & Monteith, K. L. The extra mile: ungulate migration distance alters the use of seasonal range and exposure to anthropogenic risk. Ecosphere 7 (2016) doi:10.1002/ecs2.1534.
5. Cook, R. C. et al. Revisions of rump fat and body scoring indices for deer, elk, and moose. Journal of Wildlife Management 74, 880–896 (2010).
6. Monteith, K. L. et al. Risk-sensitive allocation in seasonal dynamics of fat and protein reserves in a long-lived mammal. Journal of Animal Ecology 82, 377–388 (2013).
7. Rolandsen, C. M. et al. Accuracy and repeatability of moose (Alces alces) age as estimated from dental cement layers. European Journal of Wildlife Research 54, 6–14 (2008).
8. Cooper, S. M., Sieckenius, S. S. & Silva, A. L. Dentine method: Aging white-tailed deer by tooth measurements. Wildlife Society Bulletin 37, 451–457 (2013).
9. Boertje, R. D., Ellis, M. M. & Kellie, K. A. Accuracy of moose age determinations from canine and incisor cementum annuli. Wildlife Society Bulletin 39, 383–389 (2015).
10. Bunnefeld, N. A model-driven approach to quantify migration patterns: individual, regional and yearly differences. Journal of Animal Ecology 80, 466–476 (2011).
11. Worton, B. J. Kernel methods for estimating the utilization distribution in home-range studies. Ecology 70, 164–168 (1989).
12. Aikens, E. O. et al. The greenscape shapes surfing of resource waves in a large migratory herbivore. Ecology Letters 20, 741–750 (2017).
13. Middleton, A .D. et al. Green-wave surfing increases fat gain in a migratory ungulate. Oikos 127, 1060–1068 (2018).