In the 1960s widening U.S. Highway 27 just north of Tallahassee cut Florida's Lake Jackson into two sections. When water levels fell too low in either part, thousands of turtles, frogs, snakes and alligators would hit the road to head for the other side—where cars and trucks often hit the animals. In February of 2000 Matt Aresco, then a PhD student at The Florida State University in Tallahassee, drove through and was stunned at the sight of dozens of crushed turtles. For the next five weeks he patrolled the road between the lakes, once counting 343 dead turtles in 10 days. "It was so heartbreaking to see dozens of turtles, animals that could be 50 or 60 years old, smashed before they make it two feet onto the road," he says.

Using photos he snapped of the carnage, Aresco convinced the state transportation department to provide nylon fencing, which he set up that April along 1,200 meters on either side of the highway. Between April and August, his makeshift fence intercepted nearly 5,000 turtles that otherwise may have ended up as roadkill.

Across the U.S. vehicles hit an estimated one million to two million animals every year, the equivalent of a collision every 26 seconds, according to insurance industry records. But official numbers of animal–vehicle crashes include only reported collisions, which generally means those with large animals and that result in disabled vehicles, says Jon Beckmann, a biologist with the New York City–based Wildlife Conservation Society. "If you run over a raccoon or skunk, those are rarely reported. When you include those smaller animals, the numbers are probably well up in the millions more."

The animal generally comes out on the losing end of the encounter. Aresco calculated that in 2001 a turtle attempting to cross U.S. 27 had a 2 percent chance of surviving. He even witnessed a turtle shot through the air like a hockey puck after being struck, and says there are documented cases of these shelled projectiles going through windshields. And, according to the Federal Highway Administration (FHA), death by car represents a serious threat to 21 endangered or threatened species, including Key deer, bighorn sheep, ocelot, red wolves, desert tortoises, American crocodiles and Florida panthers. Nine panthers were killed by vehicles in 2011, 16 in 2012 and five as of April 26 this year, says Darrell Land, leader of the Florida Fish and Wildlife Conservation Commission's panther team. Those numbers represent significant losses for a population of only 100 to 160 animals that continues to lose habitat to development.

Image: Courtesy of Matthew J. Aresco.

Vehicles, according to Beckmann, Aresco and other scientists, may well be one of the biggest threats to U.S. wildlife populations. According to the Insurance Information Institute, collisions between 2008 and 2010 were more than 20 percent higher than the previous five years. And people in the vehicles suffer consequences, too. More than 90 percent of collisions with deer and nearly 100 percent of those with larger elk and moose cause damage to the car or truck involved, the FHA reports. The Insurance Information Institute estimates that collisions with deer alone cause about 200 human fatalities each year, plus tens of thousands of injuries and $3.6 billion in vehicle damage. The number of deer-related claims paid out by just one insurance firm—State Farm—increased nearly 8 percent while all other claims declined by more than 8 percent. And bad things can happen even when there's no actual collision: Drivers swerving to avoid an animal can run into each other, or off the road. Solutions do exist, although those fighting to protect animals from death by vehicle have found that putting them into practice often proves challenging.

Safer crossings
Designing roads with wildlife in mind in the first place would make the biggest difference. "Highway 27 should have been built as a bridge, for example," Aresco says. "It's easier to plan a new road, perhaps put it in a better location or build bridges or wildlife crossings, from the start."

But that strategy, of course, does nothing to improve existing roads or the larger problem of how those roads affect habitats. The U.S. highway system includes almost 6.5 million kilometers of road, not counting dirt or county ones. Those kilometers honeycomb even the most remote parts of the country, fragmenting habitat and making it hard for animals to avoid encounters with pavement. "Ungulates such as deer, pronghorn, moose and elk migrate seasonally, and roads represent a real challenge," Beckmann explains. "And carnivores move across huge areas—grizzly bears, mountain lions and wolverines can have home ranges of several hundred square miles, which gives them a high probability of crossing roads." Reptiles and amphibians, like those turtles crossing Florida's U.S. 27, also travel annually from hibernation to breeding sites.

On those 6.5 million existing kilometers the next best thing is providing a safe way for animals to get from one side to the other. Specially designed crossings, typically vegetation-rich over- or under-passes, have proved remarkably successful at reducing animal–vehicle collisions. But proper design matters.

The makeshift barrier Aresco put up in April 2000 funneled Lake Jackson wildlife through a single 3.6-meter drainage culvert. This often required animals to travel relatively long distances, putting them at risk of overheating and predation. Aresco continued to walk the barrier daily until August 2006, ferrying turtles, snakes and frogs across the busy, four-lane road—first in one direction, then the other, in response to water conditions. In those four years he recorded more than 11,000 animals, including 9,000-plus turtles, attempting to cross. Hundreds continued to die doing so; some climbed over the barriers, and the fencing material degraded rapidly in the hot Florida sun and was repeatedly damaged by mowers, vandals and storm runoff. The biologist once broke his hand making repairs, sporting a cast for six weeks.

Image: Courtesy of Matthew J. Aresco.

Aresco saw the need for a more enduring solution, and set out to find one. He formed the nonprofit Lake Jackson Ecopassage Alliance in 2002, and the group convinced the Florida Department of Transportation (DOT) to conduct a feasibility study. Completed in 2004, it recommended three to four additional culverts and permanent barriers on both sides of the highway. The local transportation authority approved the recommendation, but use of federal funds required an environmental study, a process that stretched into early 2007. Meanwhile, Hurricane Dennis dumped nearly 23 centimeters of rain in July 2005, damaging much of the still temporary fencing and resulting in a car killing a two-meter-long alligator. Rain broke fence sections again in September 2006. By then Aresco had gone to work as director of Nokuse Plantation, a private wildlife preserve two hours away, and had stopped his daily patrols. "That was really hard, knowing there were times there would be breaks in the fence and turtles killed as a result," he recalls. "But it just convinced me even more there needed to be a permanent solution, not just one person to maintain these flimsy fences for the rest of time."

Roadkill to nil
Aresco continued to face setbacks. While he and Alliance volunteers labored to replace the entire length of fence in September 2007, a passing motorist stole eight rolls of the new, UV-resistant material. Nearly 45 centimeters of rain from Tropical Storm Fay damaged the barrier yet again in 2008. Finally, under increasing public pressure, the regional transportation authority and then the DOT made the passage a priority. Construction began in September 2009, and one year later, 1.2-meter-high plastic walls directed wildlife into four culverts along 1.6 kilometers of U.S. 27. The cost: only $3 million; the result: turtle roadkill dropped to zero. That was an "extremely satisfying" end to 10 years of hard work for Aresco. "From the first day I went out there and discovered the problem, putting an end to it motivated me, no matter how long it took."

Elsewhere, wildlife crossings have achieved similar success. Vegetated overpasses spanning major interstates in Canada have reduced vehicle collisions with mule deer and elk by more than 90 percent, and in southern Florida underpasses and barrier fences have reduced panther mortality to almost zero along about 110 kilometers of road. "We know overpasses and underpasses are highly effective when done right," Beckmann says. The trade-off is that crossings can be expensive.

Less expensive options include reduced speed limits and "animal crossing" warning signs. But Beckmann says data show that speed zones don't reduce collisions, at least not without increased enforcement. Warning signs may at least increase awareness, but again, without other kinds of mitigation, probably aren't very effective, he says. For example, despite "panther crossing" signs and reduced nighttime speed limits in areas without crossings, vehicle strikes remain the number-one human-related cause of death for Florida panthers. In Nevada "bear crossing" signs installed in the Lake Tahoe area had no apparent effect on the annual number of vehicle–bear collisions, which have increased almost 20-fold since monitoring began in the mid-1990s—primarily, Beckmann believes, due to the allure of human garbage, more people and traffic moving into bear country, and an increase in the overall bear population.

Roadside animal detection (RAD) systems—computerized signs with flashing lights tripped when an animal breaks a light beam—cost more than simple warning signs but less than most crossings. These can reduce collisions with large animals, the highway administration reports, but wind may trigger false warnings, animals sometimes fail to set off the lights and the trips aren't currently designed for smaller animals. "There is a lot of potential for this technology," Beckmann says, "but reliability has been the limiting factor so far."

In 2012 RADs set to detect panthers went up along two particularly deadly kilometers of U.S. 41 in Florida's Big Cypress National Preserve, the first such system used for animals besides larger deer, elk and moose. The University of Central Florida Department of Biology launched a two-year monitoring study in 2012 to determine the system's effectiveness in detecting and protecting panthers. Researchers survey drivers, look for animal tracks where system logs show a trigger and intentionally trigger the sensors to see whether drivers slow down, according to U.C.F. biologist Daniel Smith.

Despite some successes in reducing mortality in specific areas, millions of kilometers of road remain where animals have no protection at all. Until they do, there simply aren't enough Arescos to go around.