Why does a species not adapt to an ever-wider range of conditions, gradually expanding its geographical range? In their paper published in PNAS (Proceedings of the National Academy of Sciences), Jitka Polechova and Nick Barton at the Institute of Science and Technology Austria (IST Austria) seek to answer this question, which lies at the interface between ecology and evolution.

The theory presented by Polechova and Barton suggests that any natural population is liable to form sharp margins.

J.B.S. Haldane, one of the founders of population genetics, studied this question in the 1950s. He suggested that in a habitat in which the environment subtly differs, the migration of individuals would bring alleles that are advantageous in the habitat's center to the margin.

These replace alleles that would be more advantageous at the margin, and so prevent the species' from adapting to the marginal environment. When the species is no longer able to inhabit the slightly differing neighboring area, a sharp margin to the species' range emerges.

Jitka Polechova and Nick Barton investigated the problem for populations in which genetic variance can evolve. Combining mathematical analysis with simulations, Polechova and Barton explain how in finite natural populations, sharp range limits arise even when the environment varies smoothly.

The researchers show that random fluctuations in gene frequencies (which arise in any finite population) can generate sharp margins by reducing genetic variance below the level that would be needed for the species to adapt to spatially variable conditions.

Two parameters describe the threshold at which adaptation fails. The first key parameter describes how conditions change across space, measuring the loss of fitness due to dispersal across environments.

The second key parameter describes how effective selection is relative to random fluctuations. Even when the environment doesn't change abruptly, the theory predicts that a sharp range margin may form. The new theory presented is also relevant to biological conservation.

It shows that as random processes erode adaptation, a species can abruptly go extinct in much of its former habitat, well before demographic stochasticity becomes important. A gradual worsening of environmental conditions may cause the species' range to suddenly fragment, as a single species can no longer adapt to the wide span of conditions that it encounters.