Evolution of leapfrog migration: A test of competition-based hypotheses.

Leapfrog migration is a common migration pattern in birds where the breeding and wintering latitudes between populations are in reversed latitudinal sequence. Competition for wintering and breeding sites has been suggested to be an ultimate factor, and several competitor-based hypotheses have been proposed to explain this pattern. If wintering sites close to the breeding sites are favored, competitive exclusion could force subdominant individuals to winter further away. Competitive exclusion could be mediated either through body size or by prior occupancy. The alternative "spring predictability" hypothesis assumes competition for sufficiently close wintering areas, allowing the birds to use autocorrelated weather cues to optimally time spring migration departure. To test predictions and assumptions of these hypotheses, we combined morphometrics, migration, and weather data from four populations of common ringed plover breeding along a latitudinal (56-68° N) and climatic gradient (temperate to Arctic). Critical for our evaluation was that two populations were breeding on the same latitude in subarctic Sweden with the same distance to the closest potential wintering site, but differed in breeding phenology, and wintered in West Africa and Europe, respectively. Thus, while breeding on the same latitude, their winter distribution overlapped with that of an Arctic and temperate population. Body size was largest within the temperate population, but there was no size difference between the two subarctic populations. Populations wintering in Europe arrived there before populations wintering in Africa. The largest variation in the arrival of meteorological spring occurred at the temperate breeding site, while there was almost no difference among the other sites. In general, temperatures at the northernmost wintering area correlated well with each breeding site prior to breeding site-specific spring arrival. Based on these observations, we conclude that competitive exclusion through body-size-related dominance cannot explain leapfrog migration. Furthermore, the assumptions on which the "spring predictability" hypothesis is based did not match the observed wintering ranges either. However, we could not reject the hypothesis that competitive exclusion mediated by prior occupancy in the wintering area could lead to leapfrog migration, and therefore, this hypothesis should be retained as working hypothesis for further work.