Morphological evolution is accelerated among island mammals

Abstract

Dramatic evolutionary changes occur in species isolated on islands, but it is not known if the rate of evolution is accelerated on islands relative to the mainland. Based on an extensive review of the literature, I used the fossil record combined with data from living species to test the hypothesis of an accelerated morphological evolution among island mammals. I demonstrate that rates of morphological evolution are significantly greater--up to a factor of 3.1--for islands than for mainland mammal populations. The tendency for faster evolution on islands holds over relatively short time scales--from a few decades up to several thousands of years--but not over larger ones--up to 12 million y. These analyses form the first empirical test of the long held supposition of accelerated evolution among island mammals. Moreover, this result shows that mammal species have the intrinsic capacity to evolve faster when confronted with a rapid change in their environment. This finding is relevant to our understanding of species' responses to isolation and destruction of natural habitats within the current context of rapid climate warming.

Figure 1. Distribution of Evolutionary Rate (darwins) Is Log-Normal for the Mean Values Dataset

The distribution of evolutionary rates on a logarithmic scale is shown in the inset.

Figure 2. Relation between Evolutionary Rates (Mean Values, darwins) and the Time Interval (Million Years) over which They Were Calculated

Filled circles: islands, open circles: mainland; the regression line for islands (solid line) is above the line for the mainland (dotted line). The 95% confidence intervals of the two regression lines do not overlap between the two groups at the smallest time intervals. The difference in elevation (i.e., the rate difference) between the two lines is statistically significant below 0.05 million years.

Figure 3. Effect of Geographic Variation on the Estimation of Evolutionary Rate (Measured Rate) when Populations from Different Localities Are Used

The measured rate is the product of geographic variation (x-axis) across populations 1 and 2, and local evolution through time (y-axis) within each population. The effect of geographic variation can either increase (A) or decrease (B) the evolutionary rate estimate. The size of the square is proportional to the size of the studied character. Filled square: fossil sample preserved, open square: fossil sample not preserved.

Figure 4. Evolution of the Size of a Morphological Character (Solid Line) in a Hypothetical Population

The character size of the ancestor population from the mainland increases by a small amount on the mainland. After the isolation of the population, there is a large and rapid increase of the size of this character, and the evolutionary rate (dotted line) for this character also increases. The rate of evolution on the island then decreases to values comparable to the rate values for the mainland population.