Observations of the migration patterns of Norwegian red deer show that some animals ride waves of greener vegetation as spring spreads across the landscape, whereas others jump ahead in anticipation of this higher-quality food.
The migration of animals en masse provides one of the great mysteries of life. Birds do it, bugs do it, even fish in the deep blue sea do it. But what is it that compels them to pick up and move with monotonous regularity, giving up one seemingly good home in search of another? Writing in American Naturalist, Bischof et al.1 present a field study of red deer (Cervus elaphus) (Fig. 1) in Norway that offers a tantalizing hint as to the method behind this phenomenon. Their observations suggest that migration allows deer to take advantage of temporally changing feeding conditions in different parts of the country that are dictated by variation in elevation and latitude.
Such a behavioural strategy would only make sense in a world in which growing conditions are never identical from one place to another. As every gardener knows, this is indeed the case. For example, the onset of the growing season is slower at high latitudes than it is nearer the equator, and plants at high elevation experience spring much later than those at sea level2. In a country such as Norway, these geographic realities translate into waves of vegetation 'green-up' that commence at the end of winter and gradually spread northwards and upwards to higher elevations throughout the spring and summer2,3.
Herbivorous mammals benefit most from feeding on younger, rapidly growing shoots rather than older plants, which are laden with indigestible compounds such as lignin and cellulose4. As a result, animals trying to restore weight after an arduous winter, or those shouldering the demands of lactation, could benefit by finding the crest of these green waves as they move across the landscape. Recent studies suggest that some migratory herbivores, such as elk in the Rocky Mountains3 or gazelles in Mongolia5, obtain better food than their non-migratory brethren. But what is not clear from previous studies is whether migrant animals simply 'surf' on the peak of a single green wave, continually moving with it over the course of the growing season, or if they 'jump' between peaks (Fig. 2). Because migrations often occur over vast spaces, it is difficult to measure the success of one movement strategy against other options. Bischof and colleagues solved this problem using an elegant form of statistical book-keeping that allowed them to better understand how migrating animals value different spatial locations and to assess how much of an improvement in food intake is achieved by different migration styles.
To tackle this problem, the researchers took advantage of a satellite-derived vegetation index termed NDVI, which is a remote-sensed indicator of plant abundance2. Using bi-weekly snapshots of NDVI collected over a period of nine years, they estimated the date of fastest plant growth at each location. Then, for each of 294 study animals, they prepared a matrix that linked plant-growth data with the temporal sequence of the location of the deer. Summed values along the diagonal of each matrix provide a measurement of the quality of food experienced by a given animal over the course of the growing season. The sum of other row and column combinations reflect the food value that would have been obtained with alternative departure dates the deer might have taken. If a particular deer is truly surfing the green wave, then the sum of food qualities along its matrix diagonal should exceed that of any other departure schedule.
Bischof and colleagues differentiated migratory from resident individuals. Their results demonstrate that migratory individuals perform better than resident individuals, who in turn perform better than hypothetical migratory animals would if they had lingered in their winter home ranges rather than migrating. The authors' comparisons of deer from different regions of Norway suggest that the proportion of migratory individuals scales with the potential gains that are obtainable. Surprisingly, however, they found that most migrant animals do not surf the green wave but rather jump it, arriving in their summer range well before the arrival of optimal feeding conditions. Furthermore, it seems that some animals leap from peak to peak over the course of the season. This suggests that many deer do not follow the optimal surfing strategy, but nonetheless benefit to a lesser degree by periodically jumping into greener pastures. Why they do so remains an open question — perhaps their jumping strategy stems from the need to balance feeding with physiological costs, predation risk or other constraints that might affect their survival during migration3.
As with all new approaches, there are questionable assumptions in this study. For instance, one would like to have more evidence that the date of fastest plant growth actually translates into maximum food quality. Also, there is some uncertainty surrounding the authors' classifications, because distinguishing between migrants and residents is often troublesome in species that occupy large home ranges, particularly when movements occur as a series of jumps rather than a smooth progression. Perhaps most important is the issue that, although the authors' method allows rigorous assessment of the ecological consequences of different patterns of movement timing along a single trajectory, it cannot tell us the relative value of a particular spatial trajectory against the infinite number of other conceivable patterns. This challenge invokes the well-known combinatorial problem referred to as the travelling salesman dilemma6. Although such complex spatial problems currently cannot be solved exactly, new statistical modelling techniques based on likelihood assessments could help to provide robust estimates7.
These complications aside, Bischof and colleagues' study offers a fresh take on movement ecology, a burgeoning field that blends developments in movement modelling with technological advances made available by satellite-based information systems8. The logic used by the authors should, in principle, be applicable to a broad range of animal movement patterns, ranging from nomadism to territoriality. Perhaps the day is not far off when the complex pattern of steps taken by an individual over its entire life can be predicted on the basis of the continually shifting mosaic of resources and costs with which it is confronted.
Bischof, R. et al. Am. Nat. 180, 407–424 (2012).
Pettorelli, N. et al. Trends Ecol. Evol. 20, 503–510 (2005).
Hebblewhite, M., Merrill, E. & McDermid, G. Ecol. Monogr. 78, 141–166 (2008).
Langvatn, R. & Hanley, T. A. Oecologia 95, 164–170 (1993).
Mueller, T. et al. J. Appl. Ecol. 45, 649–658 (2008).
Anderson, D. J. Theor. Popul. Biol. 24, 145–159 (1983).
Patterson, T. A. et al. Trends Ecol. Evol. 23, 87–94 (2008).
Nathan, R. et al. Proc. Natl Acad. Sci. USA 105, 19052–19059 (2008).
About this article
Seasonal sea ice dynamics drive movement and migration of juvenile bearded seals Erignathus barbatus
Marine Ecology Progress Series (2018)
Biological Reviews (2018)
Natural Resource Modeling (2017)