The birds that were being killed, further analysis revealed, weren't representative of the rest of the population. On average, they had longer wings. In 2012, for example, the average cliff swallow in the population had a 106-millimeter wingspan, whereas the average swallow killed on the road had a 112-millimeter wingspan.
"Probably the most important effect of a shorter wing is that it allows the birds to turn more quickly," says Charles Brown. Previous studies on the dynamics of flight have illustrated the benefits of short wings for birds that perform many pivots and rolls during flying and shown that shorter wings also may allow the birds to take off faster from the ground, he adds.
When the researchers analyzed the average wing length of the living birds in the population, they discovered that it had become shorter over time, from 111 millimeters in 1982 to the 106 millimeter average in 2012. The data suggested to the Browns that roadkill deaths were a major force driving this selection. Birds with longer wings would be more likely to be killed by vehicles and less likely to reproduce, the team reported online Monday in Current Biology.
The data illustrate a "beautiful trend that never could have been predicted," says evolutionary biologist John Hoogland of the University of Maryland Center for Environmental Science in Frostburg, who was not involved in the study. "We humans, because we're changing the environment so much, are adding a new kind of natural selection to these animal populations."
Few studies have looked at long-term changes in roadkill numbers, Charles Brown says, so more work is needed to determine whether similar trends hold for swallows in other areas, for other types of birds, or for mammals. "I would think that this would be a pattern that certainly might apply to other species," he says. "But there's almost nothing in the literature on historical trends in roadkills, because surveys typically last a season or two, not an extended period of years."
The new findings could also apply to birds killed by wind turbines, Hoogland adds, and they illustrate the payoff that can come with careful data collection and observation. "I think the most important lesson from this research is the paramount importance of collecting data even when you're not sure what it means or how it could lead to findings in the future."