Predicting Future Evolution (Fall 2016)

One of the activities that I regularly have my students complete in my Evolution course is called “Future Evolution“. The activity sends students on what most evolutionary biologists consider a fool’s errand: to try to predict the future evolution of some particular trait in some particular species. Making such predictions is really difficult for these basic reasons:

• So much of evolutionary change relies on random mutations, and predicting where or how new mutations might arise is nearly impossible;
• All future genetic changes will occur in the context of the existing genetic architecture of each organism, and although we are getting better at understanding this architecture we are a long ways off from being able to predict what’s possible to change in traits by making genetic changes; and
• Although there are some very clear environmental changes occurring now — most of them caused by human activities — it is difficult to know if these changes will be sustained for long enough to lead to changes in the traits of particular species.

Why send students off to make predictions when most evolutionary biologists would be loathe to make such predictions themselves? Well, as a thought exercise making these predictions is actually really valuable.

A skill that I try to teach to all of my students is to “tell an evolutionary story”. I want students to be able to explain what kinds of changes to existing traits would be required for a new trait to evolve, how the resulting trait variant might provide advantage in a particular environment and therefore increase in prevalence due to some form of selection, and explain how that overall evolutionary process would produce this new trait as an adaptation of a particular species. When we look retrospectively at traits that have already evolved, the goal is to verbalize a reasonable scenario under which this trait evolved. When we make predictions about the evolution of novel traits, the goal is to predict the evolution of a trait in a particular species that feasibly might occur given what we know about recent changes to some aspect of that species’ environment. The prediction doesn’t necessarily have to be likely to come true — even most reasonable evolutionary predictions are still low-probability outcomes — it just has to be feasible. Making these feasible predictions requires a strong understanding of how evolution works, which is why we do this exercise in our final class session.

Every semester I get some really fun, interesting, provocative, and off-the-wall predictions. Below are some highlights from two sections of this semester’s class…

Prediction: Humans will develop increased thumb dexterity due to the ubiquitous use of smart phones.

Although students are predominantly analyzing the predictions of their peers, I have to seed this activity with a first prediction to be analyzed by the first group to finish their first prediction… and this is a common seed that I use. Interestingly, this semester I got basically the same prediction from one student group. This semester three groups had to analyze this prediction, and to varying degrees they struggled with it. One of the temptations of this prediction is that the “adaptive need” that it presents seems very near to us: we all use our phones every day, and they seem like essential tools to “survival” in the modern environment. But figuring out whether or not this prediction is feasible requires really interrogating what it means to survive in a modern, technologically-complex human society. One group decided this was a feasible prediction, and the flimsiness of their explanation (“Technology… becomes more ubiquitous in our culture and humans must adapt by having more flexibility to keep up with fast-moving communication”) is a good indication that this probably is not a feasible prediction. If you can’t specifically tie a particular environmental factor to survival (as in dying or not dying) and/or reproduction (as in having offspring or not having offspring) then that factor is probably not going to influence the evolutionary process. The other two groups identified this prediction as not feasible, but to varying degrees struggled with an explanation as to why it is unlikely that our thumb dexterity will evolve in response to our technologies. The strongest explanation came from a group that explained “there are already tasks that have similar movement of [the] thumb in human’s daily activity”. This implied that we are already capable of learning how to use a variety of devices and therefore don’t actually gain any advantage by having increased thumb dexterity. To this explanation I would add that these devices are being designed to work within human constraints, and that they get better and better at working within those constraints: the cultural evolution of the smart phone makes the evolution of the thumb unnecessary. The other group that correctly identified this prediction as not feasible suggested that “the dexterity of the thumbs would not aid in the survival of the species”. This is an explanation that makes me cringe for a couple of reasons. The first is that it implies that somehow evolution is primarily a process that produces viable species, and that’s a bummer because we have spent a lot of time in class highlighting cases where evolution clearly produces what’s good for the individual rather than the species. The second problem with this explanation is that it doesn’t actually connect the processes that lead to “survival of the species” to cell phone use. And I would say that making this connection is the best way to refute this prediction: there’s no clear indication that using a cell phone with great thumb dexterity leads to better survival or reproduction. I don’t think that fast texters are less likely to die young or more likely to attract a mate. The other problem is that the environment of technology shifts so fast: even if fast texters were able to score more dates and make more babies, it seems unlikely that their offspring would also grow up in an environment that required the same old use of thumbs to manipulate communication devices. Technology evolves so fast that we just cannot know what the technological environment of future generations will look like, and that means that we probably can’t predict how these technologies will impact human biological evolution.

Prediction: The fur color of dogs will increase in variety due to exposure to chemicals in grooming materials.

Ugh, this one is cringe-worthy. One of the things I like about this activity is that it really challenges students to demonstrate that they understand the evolutionary process. But the risk is that they will demonstrate that they really don’t understand the evolutionary process. All I can hope is that there is some pranksterism behind this prediction, because it is pretty much completely off. First of all, I don’t know of any evidence that grooming materials contain chemicals that change the fur color of dogs. Any product that did would likely be sued out of existence by indignant dog owners. Second, even if the products did cause changes in fur color, those changes would not be heritable. Assuming for a moment that somehow these chemicals are mutagenic (again, a wild assumption) and that mutations cause the hair color change, we have to realize that hair follicle cells would have to be the ones mutated. And we don’t pass on our somatic cells, we only pass on our gametes. So this prediction is silly, and definitely not feasible.

Prediction: Dog tails will become shorter due to humans clipping dog tails for aesthetic reasons.

Another cringe-worthy Lamarckian prediction, at least on the face of it. Obviously clipping a dog’s tail does not remove the genes for developing a tail from its gametes, so there’s no way that this acquired characteristic could be passed on. Perhaps it is possible — as can be seen in some breeds with tails that biologically develop to be shorter — that humans might breed dogs to have shorter tails, but in this case the cause of change would human selective breeding practices, not the act of clipping dogs’ tails.

Prediction: Humans will develop an increased propensity to dance due to the popularity of fad dance moves.

Whether or not humans select mates based on their ability to dance is still unknown; the one study that purported to show that dancing ability was a sexually-selected trait in humans was later shown to be based on fraudulent data. But this is a reasonable prediction based on how importance dance is as a social activity cross-culturally: humans do appear to love to dance, and perhaps that’s because dancing is a way of displaying one’s fitness. But there are a couple of facets of this prediction that probably make it not feasible. The first is the use of the word “fad” in the prediction. If these fad dance moves are really coming and going rapidly, that means that they can’t be a stable component of the environment for long enough to really affect mating success. I suppose one could argue that while dance moves come and go, they all share a common need for dexterity, rhythm, and coordination. But what makes me skeptical about this prediction more generally is that I don’t believe that modern-day humans make “dance ability” a primary sexually selected traits. With so many other characteristics to consider in our potential mates, it’s hard to imagine that there’s very strong sexual selection on dancing ability. But maybe I am just biased by being a father of three who can’t dance very well.

Prediction: Orangutans will become bipedal and lose their tails due to deforestation and loss of their food sources.

This one is another head-scratcher. Thank goodness this is not a class that emphasizes natural history or primatology, because the last time I checked orangutans don’t have tails (like all their relatives in the Hominidae). But even if you lose the tail part, this prediction doesn’t make sense. Where are these orangutans going with their new bipedal trait? What environment would they be adapting to, even if such rapid evolution of bipedalism were possible? The sad reality is that species like orangutans are highly unlikely to evolve to the rapidly-changing environments they now face as they come into closer contact with humans and human activities. Human cultural change happens so fast that we probably can’t biologically evolve in response to these changes, and neither can our animal comrades. There’s a kind of danger to this prediction in that it ignores a well-understood reality: if we don’t preserve the habitats of species like orangutans, they will go extinct. Evolution can’t save most species facing human threats, because human threats cause change that is too rapid.

Prediction: Polar bears will become better adapted to an aquatic lifestyle, eventually becoming whale-like species, in response to ice caps melting.

Again, this prediction has a pace problem. Many, many mutations must arise and be selected for in order to transform current-day polar bears — who are good swimmers but certainly more adapted for land — into a species that can survive in the water alone. There’s just no way that such a complex evolutionary process could unfold on the incredibly short time scale of climate change. Again, evolution is not likely to fix the biodiversity problems created by human impacts, at least not on a time scale that matters.

Prediction: Humans will be less inclined to pass on their genes due to lower libido, caused by financial burden of having children and career aspirations.

Wow, this prediction is near and dear to my heart because it relates directly to my book-in-progress Breeders, Propagators, & Creators. This is clearly a prediction that is not feasible because it violates one of the chief tenets of biological evolution, which is that in order for a trait to evolve it must be heritable. If people who are “less inclined to pass on their genes” in the current cultural environment don’t have offspring, their inclination is being selected out of the population. And that actually implies that — if anything changes at all — in the future people will be more resistant to the temptation to prioritize career over parenthood. Of course this all assumes that the propensity to have children in ‘difficult’ or ‘culturally distracting’ environments is influenced by genes; although I don’t yet know of any evidence suggesting that motivation to become a parent is influenced by genetic variation, it would not surprise me if it was. But clearly there’s a very strong cultural component to the decision to reproduce: tell me something about the socioeconomic and cultural background of a person and I am going to be much more likely to be able to predict their likelihood of having children. This is an intriguing prediction because it asks us to ask what effect all this contemporary not having of children will have on the future of the human species, but I can’t see how our inclination to have children will biological evolve to decline.

Prediction: Seagulls will evolve to become less dependent on ocean food sources due to the wide availability of human garbage as a food source.

This prediction is funny, but feasible. I would have liked to see the group that made this prediction focus more clearly on what actually evolves here: this would be a shift in behavior from focusing on searching for marine sources of food to switching to sources of human-provided food. This shift might already be happening, as we see a lot of seagull populations persisting far inland around landfills and other sources of easy human food; it would be interesting to see if anyone has tracked the movement of these populations to see if they obligately feed on human food sources. What seems most reasonable is that populations of these seagulls might be selected to better locate, consume, and digest human food sources. That suggests that a separate commensal seagull populations could arise, much in the way that rats and cockroaches have come to have their own populations in human-dominated areas. The interesting question here is whether this seagull population could become a different species. If the behavioral adaptations required to access human food are different enough from the behavioral adaptations required to access marine food sources, it is possible that there could be selection for reproductive isolation. Could such parapatric speciation happen? Maybe, but only if some behavioral barrier to interbreeding also emerged.

Prediction: Coral species will become better able to withstand changes in water temperature caused by climate change.

This is a tough one. Like the orangutan example above, the validity of this prediction rests on the pace of environmental change and the extent of biological evolution required to respond to that change. While massive dieoffs in places like the Great Barrier Reef raise fears that extinction rather than adaptation will be the fate of corals, these extreme events also are exerting strong selection for corals that do possess the genetic variations that allow for survival in warming seas. But corals have been shown to adapt to increased water temperatures, and there is reason to believe that they can adapt to more widespread warming. So I think that this is a feasible prediction, but not one that should cause us to underestimate the problem of coral reef dieoff. Even if some corals can adapt to climate change, the dieoffs resulting in this evolution will be massive and the time for these species to recover is going to be very slow. And if climate change continues unabated, at some point the limits of adaptability are going to be reached and we may lose many coral species for good.

Prediction: Human intestines will get shorter due to eating processed food.

This is a funny prediction, one that may evoke some “who cares” reaction. But I think that it is interesting to consider how something as central as our food source may influence our anatomy. If this is a feasible prediction, it will likely happen due to release from selection rather than direct, positive selection. The difference between these two processes is critical. In one process, the population drifts away from a previously-crucial trait because the trait is no longer important. I can see where humans might be released from selection for a long intestinal tract, because we no longer gather foods that require as much digestion. We also have access to food in such abundance that the efficiency of a longer digestive tract may no longer be an advantage. So just as our vision is drifting further and further away from 20/20 due to the persistant availability of corrective lenses, so too may our digestive tracts evolve away from an optimal length for extracting nutrients from unprocessed foods. But there are also scenarios in which I could see this that this prediction is not feasible. There’s not going to be any selection for a shorter digestive tract, because the only real advantage of having a shorter digestive tract would be efficiency, and our abundant food access makes that sort of efficiency irrelevant to our survival. So maybe this change can’t happen because it relies on drift rather than selection to happen. But the other question that hangs around this prediction is do most people on the earth eat processed foods? I suspect that the answer is no, and so until the adoption of processed-food diets becomes more widespread, it is a bit difficult to see this as a feasible prediction.

Prediction: Deer coats will develop to be grey “urban camo” in response to living in human-dominated environments.

Another funny one that’s maybe too easy to dismiss immediately. Deer clearly now live in and among people, and their interactions with humans are now a major source of mortality. So we have the prerequisite for evolution: a relatively novel environment that creates new challenges for survival. But is this the right prediction for this environment? The prediction supposes that the main risk to deer is that they are not sufficiently hidden in the human-dominated environment. I imagine that means that those deer who don’t blend into the urban environment are more likely to be predated by human hunters. But is that right? The last time I checked, there were plenty of deer running around in the suburbs, but they are not being picked off by rifle-wielding hunters running across pristine green lawns. A major source of their mortality appears to be collisions with vehicles. If collisions with human vehicles — rather than hunters or other animal predators — represent the biggest risk to deer living in human-dominated environments, then you might predict that these deer populations would evolve more conspicuous coats. If only the most visible deer escape being hit by cars, the population could actually evolve to be less camouflaged.

Prediction: Elephants will evolve to no longer have tusks.

This is perhaps the most feasible prediction made this semester, and that’s because this process of declining tusk size has already been observed. That said, whether this prediction’s rather extreme end result is fully realized will depend on a couple of unknowns:

1. Will ivory continue to be poached and sold in spite of being illegal where elephants live and almost everywhere where ivory might be sold?; and
2. Will elephants as a species be around long enough to evolve this radical shift in traits?

Sadly I think that the answer to question number one is probably more likely to be “yes” than question number two. The problem is that the selective process that’s driving smaller tusk size is also driving the species towards extinction. Nonetheless, this is a feasible prediction given what’s known about the effects of poaching on tusk size.

Prediction: There will be an overall decrease in the diversity of human skin color due to increased interbreeding of people of different geographic ancestry.

This was a prediction made by no less than three groups of students, and it seems feasible. Some groups got tripped up by the concept of race, which really is not a biologically-meaningful category, but I am going to treat this along the lines of geographic origin. Clearly the process of “de-regionalizing” human populations has been underway for many centuries; the popularity of shows that trace the often-complicated geographic ancestry of people alive today [1, 2] highlights both our fascination with this issue and the prevalence of gene flow between human populations of the near past. But we also live in an unprecedented time of global movement, and so the environment for breaking down regional characteristics like skin color certainly exists. One question is whether this environment will persist: although in western countries we all know people of radically-mixed regional origin, the interbreeding of people from diverse backgrounds is accentuated in developed economies. For this prediction to come true, more people globally are going to have to join this process of dispersing everywhere else, and it is hard to know if that’s going to be a trend that continues. While all of the groups making and analyzing this predictions seized on the global nature of human migration to justify this prediction, none of them focused on the original adaptive value of regional skin colors, which has to do with the tradeoff between protection from harmful UV and sufficient UV exposure to synthesize vitamin D in the skin. It also pays to recognize that the “de-diversifying” of human skin color would be as much due to release from selection as it would be from global migration. Because we now how technologies like sunscreen and vitamin D supplements, there’s no broad advantage of having a particular skin color in a particular region.

As you can see from the predictions above, my students are still — after a full semester with me — making some very questionable predictions. Some are the cringers, predictions that really make me wonder whether my students have learned the basic process of evolution. Others make more subtle errors, ones that result not from misunderstanding the evolutionary process but from misjudging its pace and constraints. It would be tempting for me to set the goal of having all these predictions be feasible in this activity, but I think that such a goal would be unrealistic. While the game we play in this activity incentivizes students to make valid predictions, many of them are probably making predictions that they know are a little crazy, perhaps to trip up the other students who have to analyze their predictions and perhaps out of the inability to make a valid predictions. As I said from the outset, evolutionary prediction is a fool’s errand, so making even feasible predictions is hard. What I focus on is what can be learned by analyzing each prediction.

To read about Predicting Future Evolution posts from other semesters, navigate here.

You can check out the instructions for this activity here and the playsheets for this activity here. Feel free to contact me if you want help in implementing this activity in your classroom!