Here we’ve had, in many areas across the country, a strong selection pressure in the form of deadly roads. Yet roadkill doesn’t seem to me to have become less (here at least, yes data point of one, await any more rigorous studies if any exist).
Which may mean a few things: that road crossing behavior is not inherently genetic, there has not been sufficient time to see any longer-term effects, there may be cognitive constraints on the part of the species in question, or they may indeed think (at some level) that whatever is on the other side is worth the risk (so for every dead animal a hundred may have crossed that night safely).
While I do see dead squirrels with regularity, I also see them often using power lines to cross a busy road.
In my opinion, this is the most important factor. The number of individuals killed is relatively small compared to the total population of the species in the area, and the odds of any one individual being killed is small compared to the number that successfully cross the road. And the mortality rate is probable small compared to other factors such as predators, disease, lack of food, etc.
In areas where there are a lot of roads, the benefits of crossing the road may exceed that of having to restrict activity to the smaller areas contained within roads. In areas with few roads, the population at risk from roads is much smaller than that that doesn’t experience roads, so any selective effect on the former will be swamped by the latter.
Cars come on far faster than most predators, so most species aren’t adapted to react fast enough. Extensive road systems with fast traffic are less than a century old, which probably isn’t enough time to force adaptations to your predator avoidance system.
To get to the other side. Too bad it didn’t work out.
Sounds like a poor joke, but this is GQ, so I wouldn’t joke.
“Crossing a road” isn’t a specific behavior. It’s just (for land animals) “walking.” Presumably, in the course of a survival-oriented activity, like foraging for food.
So, what’s evolution going to make the animal do? Not walk? Good way to overgraze one spot, let alone never ever meet a member of the opposite sex and reproduce.
Recognizing an unnatural threat is also unlikely. And frankly, the reason some critters wind up as roadkill is because a valid survival reflex (in response to “natural” hazards) is counterproductive against cars. Armadillos jump when startled. Probably helpful when what’s startling you is a snake or a coyote; bad when what startles you is the car passing over you. Likewise, a deer freezing in the headlights makes fine sense if the “headlights” is the eyeshine of a night-hunting predator.
Because not enough animals in a species get killed for natural-selection to really play a role.
Suppose that every year, 5% of opossums get killed by oncoming traffic. The 95% can still reproduce just fine and, given the plus-or-minus nature of reproduction (maybe more babies in the litter one year, less the next,) plus the other variables like food supply. climate, predators, etc., the car killing really does not make a significant impact.
Roadways are so pervasive, and the speed of vehicles is so fast compared to the visual acuity and instinctual response of most creatures that it is essentially impossible to select out characteristics that cause animals to cross roadways in front of vehicles. Many roadkill animals are also nocturnal or crepescular and may be stunned into immobility by the sight of approaching bright headlights. Scavenger animals may also prey on roadkill before becoming victims themselves; anyone who has driven through Joshua Tree at twilight and seen the endless string of reflections from coyotes waiting for a spooked rabbit to be struck can attest to this.
Of course, it isn’t just wild animals that are subject to this. Being struck by a vehicle is the most common traumatic death for outdoor domestic pets such as cats and dogs, and ~130k people are hit by cars in the United States every year with over 5k fatalities. Rates are similar in European nations and even more common in developing nations with less traffic control and driving rules enforcement.
High speed 50+ kph) vehicles have only been common for about the last hundred years, so any evolutionary pressure to evolve characteristics to protect against exposing a creature to being struck had not had a long time in evolutionary sense to develop, but even then for many animals the extreme differnence in speed and context of automobiles (and aircraft, as bird strikes on airplanes are also very common) make it difficult to conceive of adaptations in behavior or senses that would protect them. People avoid being struck by following rules such as walking on the sidewalk and crossing at designated intersections in the hope that drivers are attentive and follow rules (which is not always the case) but animals are generally not trained to understand these rules and expose themselves to hazards that they cannot comprehend.
And a very great many animals do, in fact, respond appropriately to cars. Consider, for instance, all of the times that you’ve almost hit a squirrel or a pigeon: The key word there is “almost”. Now count up how many times you’ve actually hit one: It’s a lot less, isn’t it? Apparently, the dodge tactics the squirrels and pigeons are using actually are pretty good.
Remind me how many thousands of years we’ve had highways wiping out these critters? Or should that be millions or hundreds, I forget the timeframe for evolutionary selection for animals of the roadkilley size and lifespan…
I know fruit flies and influenza viruses adapt alarmingly fast, and there were those squirrels with dark fur that became predominant during the height of the coal-burning era, and squirrels are kinda roadkillesque in size and character, but selecting against a characteristic that puts one at greater risk may take longer than selecting for one that better ensures survival (?)… anyway, the 80-some-odd years of automotive hegemony seems awfully short to assume it should’ve had a culling effect, don’t you think?
It was mentioned in the OP, but lately I’ve seen a LOT of squirrels around this area (deep suburbs with lots of fast ‘main roads’ nearby) crossing the streets by running up a utility pole, crossing on the cable, and either running down or jumping into a nearby tree on the other side.
If this is an adaptation among squirrels, it’s happening pretty fast, considering neither utility poles nor fast cars have been around long. But maybe it’s just a normal squirrelly behavior that happens to fit the circumstances - something like “use overhead access when available, so as to avoid ground-based preditors”.
I’ve noticed corvids, especially magpies, commonly hanging around on the shoulder of roads, a meter or so from cars, apparently totally unconcerned by high speed passing vehicles, but when they’re scavenging from roadkill in the middle of the road, they take off in plenty of time in front of oncoming cars. It certainly looks like they understand where on the road the cars drive, and where they are unlikely to be. It’s rare to see a corvid as roadkill- and the rare exceptions seem to be almost entirely juveniles. That’s probably learned behaviour, rather than evolved behaviour, but it seems to be working.
Pheasants, on the other hand, freak out when cars are moving near them regardless of whether they’re in the path or not, then half the time they fly a little further down the road and land on this nice flat surface until they see this scary car coming towards them and take off… again and again. There are a lot more roadkill pheasants around, though they don’t hang round roads nearly as much.
I suspect that the intelligence to process that roads work to fairly predictable rules is important, and that’s a big difference to evolve.
Incidentally, elephants also seem to recognise that cars stay on roads, though I admit my observations of wild elephants is a bit more limited.
That’s the likely explanation. Squirrels are not evolved to travel on utility poles and wires. But they are evolved to travel on trees and branches. So modern industry gave them something they were equipped to handle.
Deer are evolved to travel out in the middle of a clearing where no predator can sneak up of them before they have a chance to run away. And they’re evolved to escape a predator that can run around thirty miles an hour. So deer are not equipped to handle modern highway traffic.
Indeed; it’s easy imagine that the selection might work the other way in some cases. Animals too skittish to cross an unusual surface might well be outcompeted by those that benefit from a larger territory, despite the risk of death.
I’d say this is a big part of it. Deer are a fairly sizable problem in my area. Unfortunately, we make conditions favorable for deer to hang around near us by having plenty of food around, pushing the predators away, and not hunting in residential areas. So the relatively few which get turned into cherry pie by cars and trucks aren’t enough to stress the population.
How many generations would it take for changes to occur? ISTR reading that 10 was a minimum. Now, if high-speed vehicles (over 50 km/h as mentioned before) have been around for only 120 years or so, how many deer generations is that? Squirrel generations?
Also, deer have an instinctive “choose a random direction to bolt in” behavior. If they always bolted the same direction they’d be predictable. So they mix it up. Unfortunately with cars that means half the time they jump out in front of the car instead of away from the car.
At least with birds “go up” is a pretty standard strategy for evading danger, which works for cars.
Where I live roadkill is about 40% deer, 40% raccoon, and 20% other. The deer and raccoon are just because where I live it’s infested with raccoons and deer, a fact which drives our nervous dog crazy every night.
Is that by number or by mass? By number, squirrels are probably more common than deer or raccoon; they’re just not as noticeable, and they’ll be cleared away more quickly by scavengers.
The way Natural Selection works, is simple, but an amazing number of people get it wrong.
The “selection effect,” in this case, humans in cars, has to cause the death of a majority of the species before it can procreate. That’s it. And cars do not accomplish that, because crossing roads is not necessary for any species to maintain itself.
Killing “a whole lot” of a given creature only matters if the kill rate is higher than the replacement rate, over a long period of time.
Dear and squirrels obviously procreate a lot faster than the number who die under our tires.
The species that we’ve driven to extinction, or near extinction, were all either purposely hunted to try to get them all (wolves, tigers, bison and the like), or accidentally wiped out while we were doing something else (condors, honey bees, etc)
