MAKE MORE OF THEM
In 2004, a group of twelve distinguished scholars—conservation biologists, paleoecologists, mammologists, and community ecologists among them—met at Ted Turner’s Ladder Ranch in the Chihuahuan Desert of New Mexico and developed a visionary plan for North American biodiversity. They proposed to reintroduce a small number of large-bodied animals, many of them endangered, into what little wild habitat remained on the continent. In doing so, they would protect North American biodiversity from continued decline. As a bonus, some endangered species would be provided a new, safe place to live and a better shot at survival.
Their premise was simple: big animals are integral to any ecosystem. Big animals play key roles in recycling nutrients, distributing seeds, turning over soils, and knocking down trees. Big animals are, however, missing from the North American landscape, largely due to terrible things that humans have done. To restore North America to a more balanced state, it is therefore necessary to restore big animals.
The group of scholars pointed out that restoration efforts tend to focus on reestablishing the flora and fauna that were present in North America when Europeans first arrived several hundred years ago. By that time, however, most of the big animals that had dominated the landscape throughout the Pleistocene ice ages were already gone. The group proposed looking further back in time to what they believed was a more appropriate baseline for North American restoration. A better target, they insisted, would be the Late Pleistocene—before human arrival and before the megafaunal mass extinction. The Late Pleistocene, they argued, was a time during which a diverse community of large herbivores maintained a diverse community of vegetation and were preyed upon by a diverse community of large carnivores. Naturally, the continent looked very different during the Late Pleistocene than it did when the first European colonists arrived.
Restoring North America to a Late Pleistocene baseline would be challenging, especially since many of the species that dominated the landscape at that time are now extinct. Not all of them are gone, of course. Some species survived, albeit in much more diminished ranges, for example, North American bison and giant desert tortoises. These species could be reintroduced wherever suitable habitat remains within their former range. Species that have gone extinct, such as camels, horses, and mammoths, could be replaced by proxies—living species capable of filling niches that were left vacant when the megafauna disappeared. Where reasonable proxies could be found, these species could be introduced into habitats that were once occupied by their extinct evolutionary cousins.
The plan for restoration was to start small and proceed in stages. First, Bolson tortoises (also known as Mexican giant tortoises) would be reestablished across the Chihuahuan Desert, which stretches from central Mexico northward through western Texas and parts of New Mexico and Arizona. The Bolson tortoise is North America’s largest living terrestrial reptile. Although it was distributed across the Chihuahuan Desert during the Pleistocene, the Bolson tortoise is now restricted to a tiny, semiprotected refuge in north-central Mexico. Fortunately, the former range of the Bolson tortoise still includes some ideal habitat for reintroduction. Big Bend National Park in Texas, for example, used to be home to Bolson tortoises, and reintroduced tortoises could presumably get right back to the business of grazing on bunch grasses and digging burrows. It is unlikely that tortoise reintroduction would significantly alter the existing ecosystem of Big Bend National Park, other than to disturb the soil in a useful way. And it is unlikely that the tortoises would require much human intervention to survive. The most visible effect of Bolson tortoise reintroduction would probably be an increase in tourism to the park, as people realize they might be able to spot an eighty-year-old giant tortoise in its native habitat.
After the tortoise, the group planned to introduce horses, donkeys, and camels across the wilderness regions of western North America. Not just feral domestic horses and donkeys, but also their wild Eurasian cousins: the Przewalski horse and Asiatic wild ass. The group would also introduce camels—wild camels if possible, but domestic camels would suffice.
Why these species? When the ancestors of present-day horses and camels were living in North America (both horses and camels evolved in North America), woody plants were heavily grazed by large herbivores. This opened up space in which other types of plants could flourish, increasing floral biodiversity. A greater diversity of plants could sustain a greater diversity of herbivores, both large and small. And these, in turn, could support a greater diversity of carnivores. Large herbivores also act as efficient distributers of both nutrients and seeds. Their feet turn over the soil as they roam and run, their bodies transport seeds over long distances, and their excrement fertilizes the soil. Thanks in part to these animals and the roles they played within the ecosystem, Pleistocene North America was a mosaic of plant biodiversity that was capable of supporting a mosaic of animal biodiversity. Reestablishing horses and camels may help to restore this biodiversity.
Of course, the group was aware that introducing horses and camels to wild land in North America would be somewhat more controversial than introducing Bolson tortoises to desert ranches and US national parks would be. Feral horses and donkeys are considered by some people to be pests that compete with livestock. Any plan for reintroduction would have to balance the needs of the people who use the land with the potential benefits to the ecosystem. Strategies would need to be developed both to educate the public about why having these animals around might be good for the ecosystem and to teach people how to interact with these animals when they come into contact or conflict. Equally importantly, legal guidelines would be required to manage introduced populations and mitigate any potential negative consequences of reintroduction. At least some of the introduced species would not be native to North America—Bactrian camels, for example. Developing these strategies might therefore require new and creative thinking by legal scholars and wildlife managers. And finally, while Bolson tortoises probably won’t need human intervention to maintain reasonable population sizes, populations of horses, donkeys, and camels could explode if left unchecked, with potentially devastating consequences to the ecosystem that their introduction was meant to preserve. After all, during their Pleistocene heyday, large herbivores were kept in check by large carnivores that are now extinct.
Which brings us to the next stage of the plan: cheetahs and lions.
And elephants.
African cheetahs, African lions, and Asian and African elephants. In North America.
Just as Bactrian camels were proposed as proxies for the extinct North American camel, Camelops, African cheetahs would take the place of the extinct American cheetah, Miracinonyx, and African lions would fill in for the extinct North American lion, Panthera leo atrox. Asian and African elephants would fill the niche once occupied by mammoths, mastodons, and gomphotheres.
To be clear, the plan was not to take animals from Africa or Asia and bring them to North America—this was one of the many angry accusations that came in the wake of the plan’s release to the general public—but to identify and translocate animals already in captivity in North America to more realistically natural settings.
Needless to say, the plan to rewild North America did not pass quietly under the radar. Josh Donlan, who was the lead author on the two-page article1 that appeared in the journal Nature, received the bulk of the public backlash. Donlan reported a pretty even mix between lovers and haters of the plan, with responses falling mostly within the range of what was predictable. There were, however, some surprises. Among the lovers were a handful of ranchers who were thrilled that they might be able to use elephants to keep the brush on their land at bay, as elephants would be much less expensive to operate than the heavy machinery they rely on at present. These ranchers were understandably less keen on the big cats.
FACILITATED EVOLUTION
The motivations behind the rewilding movement are similar to those that underlie my interest in de-extinction. Proponents of rewilding aim to restore biodiversity to ecosystems that have been negatively affected by extinctions. They hope that rewilding, by reestablishing lost biodiversity and re-creating missing interspecies interactions, will allow a much richer, more productive, and more diverse community of plants and animals to prosper. De-extinction could do the same thing, but with one small but important difference. The plan proposed by Donlan and his colleagues to rewild North America included the introduction of Asian or African elephants. However, Asian and African elephants never lived in North America and may not be particularly well adapted to the North American climate, which is much cooler than that in which they evolved. De-extinction also aims to introduce elephants into habitats in which present-day Asian and African elephants may not survive. But de-extinction will first prepare these elephants to live in a cooler climate by resurrecting adaptations that evolved in their cold-adapted cousins—mammoths—and inserting these adaptations into the elephants’ genomes.
It is precisely in this way—by resurrecting adaptations from the past within the genomes of living organisms—that I imagine de-extinction as a powerful new tool both for biodiversity conservation and for the management of wild and semiwild habitats. Take mammoth de-extinction, for example. Some advocates for mammoth de-extinction probably don’t care what ecological role unextinct mammoths might play on the Siberian tundra. Some probably don’t even care if they ever make it to the Siberian tundra, as long as they make it to a zoo or a park where they can be observed and possibly ridden. I, however, and others including George Church and Sergey Zimov, care very much about how unextinct mammoths—or, more correctly, genetically engineered Asian elephants—might change the Siberian tundra. In fact, their potential to invigorate the Siberian tundra is precisely why we are motivated to support this project.
So what would the Siberian tundra ecosystem gain from the introduction of cold-tolerant elephants? Working within the boundaries of his Pleistocene Park over the past few years, Sergey Zimov has shown how large herbivores—bison, muskox, horses, and several species of deer—can transform a mostly barren tundra into a rich grassland over the course of only a few seasons (plate 16). It’s simple. They trample and graze the tundra, turning over the soil, dispersing seeds, and recycling nutrients. Their increased grazing stimulates the growth of grasses, which increases the density and nutrient quality of the available forage. Not all of the grass that grows can be consumed during the summer, leaving sufficient resources to support the animals during the Siberian winter. After the snow falls, the herbivores return regularly to the richest areas of grassland, trampling down the snow and eating everything beneath. Above ground, the grasses are consumed entirely. Below ground, the roots remain intact. In essence, Zimov’s research has shown that the interaction between herbivores and arctic grasslands is self-sustaining. When one part of that interaction disappears, so does the other.
Zimov believes that the Siberian tundra could be transformed into rich grasslands reminiscent of the Pleistocene steppe tundra simply by returning large herbivores to the ecosystem. Revived steppe tundra would provide resources and habitat for other endangered species, including wild horses, saiga antelopes, and Siberian tigers. Zimov argues, however, that the missing critical piece to his Pleistocene puzzle is elephant-sized. Large herbivores play different ecological roles within a community than do smaller herbivores. Large herbivores knock down trees and trample bushes, for example, and transport seeds and nutrients over much longer distances than small herbivores can.
There is another, potentially more significant benefit to having large herbivores graze the Siberian tundra. Although the uppermost layers of the Siberian soil freeze and thaw with the seasons, the soil beneath these layers remains relatively constant in temperature throughout the year. This constant temperature is roughly equal to the mean annual air temperature, with an important caveat. During the winter, ambient air temperatures in Siberia can be as low as –50˚C; however, snow sitting on top of the permafrost insulates the permafrost soils from this bitter cold, keeping them warmer than they would otherwise be during this time of year. Prior to the extinction of mammoths and other ice age megafauna, this snow would have been completely removed in some places and trampled in others, destroying its heat-insulating properties. The soil temperatures would have been dramatically colder than they are today. Although the number of grazing herbivores in Pleistocene Park is too small to have this same effect, it is nonetheless clear at smaller scales: Zimov estimates that the soil beneath grazed land in his park is somewhere between 15˚ and 20˚C colder during the winter months than that beneath ungrazed land.
Scientists estimate that there may be as much as 1,400 gigatons of carbon currently trapped in the frozen arctic soil—almost twice the amount of carbon that is in Earth’s atmosphere today. As global temperatures rise, the permafrost is melting and the carbon trapped within that permafrost is being released. If Zimov is right, then reintroducing mammoths into Siberia—or rather, introducing cold-tolerant Asian elephants into Siberia—will actually slow the accumulation of greenhouse gases in Earth’s atmosphere and therefore the rate of global warming.
Importantly, the scenario above does not require the resurrection of a mammoth. All it requires is a mammoth proxy: an elephant that’s genetically engineered to survive in Siberia.
ONE PLUS MORE MAKES A POPULATION
One elephant will not convert a denuded landscape into a flourishing and diverse ecosystem, regardless of how many genes were altered and how well-adapted the resulting animal is to living in that environment. However, this is exactly what we will have when the first phase of de-extinction—creating a living organism—is complete: one magnificent, healthy, genetically engineered elephant. Getting this far was certainly no walk in Pleistocene Park. Now we have to do it again.
To forge ahead with the second phase of de-extinction—releasing populations into the wild—we need to answer three questions. First, how many individuals will be required to establish a healthy population of our resurrected species? Second, how genetically diverse will the population need to be in order for it to be sustainable? Finally, where and how will this population be raised and nurtured so that it can eventually be released into the wild?
Several options are available to create a viable population of genetically engineered individuals. In the absence of significant improvements in the efficiency of genome editing, it is likely that only one cell will end up with all the requisite genomic changes. We could make more than one animal using this cell by growing that cell into a colony of identical cells—this is often called a cell line—and then using multiple cells from that cell line to create clones via nuclear transfer. One drawback to this approach is that all of the animals born will be genetically identical and, consequently, our population will have no genetic diversity. As another option, we could breed the engineered individuals with individuals that are not engineered. This would have the benefit of increasing the population’s genetic diversity but may result in the loss of the genetic changes that we worked so hard to engineer as nonengineered genomes are bred into the population. A third option would be to start from scratch and reengineer the genome edits into cells isolated from a different individual. This would also increase genetic diversity but might not result in an organism with the same or even the desired phenotype. Because every genome is different, and all of the genes within a genome interact with each other, edits that have the desired phenotypic result in one cell may not have the same result when interacting with the genome of a different cell.
Given how hard it will be to create even one genetically engineered individual, and given that it will be just as hard to create a second edited individual that is not a genetic clone of the first, perhaps we should take a step back and ask whether genetic diversity is actually necessary for a population to survive. Do we really need to worry about creating a genetically diverse population?
The answer is probably.
Genetic differences between individuals are the substrate for adaptive evolution. If everyone in a population has the same genotype, then everyone will also have the same, or an extremely similar, phenotype. Everyone will be equally likely to survive and to reproduce. Of course, everyone will also all be equally unlikely to survive. If a disease sweeps through the population, for example, everyone will be equally susceptible to that disease. If the environment suddenly changes—perhaps there is a severe drought and an important source of food disappears—no individual will be better able to adapt to that change in resource availability than any other individual will be. Populations with high genetic diversity are buffered against disease and environmental fluctuations. Some individuals in these populations will be more likely than others to survive and reproduce. The genetically diverse population will adapt and survive.
Are high levels of genetic diversity absolutely necessary, however? Low levels of genetic diversity have been linked to poor health, decreased reproductive success, and even physical abnormalities, such as the crooked tail that was frequently observed among Florida panthers prior to their hybridization with panthers from Texas. Some species, however, have extremely low levels of genetic diversity but little measurable consequence to their ability to survive. Polar bears, for example, have extremely low levels of genetic diversity, but they have had the same tiny amount of diversity for at least the past 100,000 years. During that time, polar bears survived two ice ages and the present warm interglacial period. Nonetheless, as the habitat to which they are very specifically adapted disappears, their lack of genetic diversity may be their downfall: the more genetically diverse a population is, the greater the chance that this diversity will recombine in new ways, resulting in phenotypes that can adapt to surviving in a different environment.
Clearly, genetic diversity and the adaptive potential that genetic diversity provides are important, and a healthy population cannot be made up entirely of genetic clones. While it won’t be the simplest solution, the most likely solution to the diversity problem will be to engineer cells taken from different individuals and use multiple cells to create a genetically diverse population. When editing the genomes within these cells, we will need to be certain that the edits are made in both chromosomes of every cell that is used. That way, the population will be genetically identical at these particular loci, and the target phenotype will be maintained even after the population is released into the wild.
While genetic diversity will be important to consider when creating our resurrected population, we must also keep in mind that diversity is not the only factor that determines whether a species is sustainable over the long term. If we were to survey genetic diversity among living primates and use this information to decide which primate species is most in need of protection, the result would shock most of us. The primate with the least amount of genetic diversity is … us. Humans have almost no genetic diversity, whereas other primates, including chimpanzees and gorillas, are doing just fine. Engineering genetically diverse populations will be important for de-extinction, but, ultimately, it will not be as important as finding a stable, healthy, and sufficiently large wild space into which our population can be released.
FROM THE BIRTH OF ONE TO THE REARING OF MANY
The second phase of de-extinction involves not only creating multiple individuals, but also rearing and nurturing these individuals, moving them out of captivity, and establishing populations in the wild. Ideally, this second phase would culminate with the establishment in the wild of multiple genetically robust, healthy, self-sustaining populations that are resilient in the face of environmental change. Phase two is certainly not going to be easier than phase one.
As a start, the babies must develop into adults. They must develop both physically and behaviorally, taking on the characteristics that they have been engineered to express. Most likely, several generations will be born and raised in captivity before a sufficient number of individuals are available to be released into the wild. Populations living in captivity, possibly for decades, need not only to survive, but they must also learn how to live. The individuals that make up these populations need to learn how to feed and protect themselves, how to interact with others, how to avoid predation, how to choose a mate, and how to provide parental care to their own offspring. Understanding how a species might fare in captivity is therefore an important consideration in deciding whether a species is a good candidate for de-extinction.
Humans have lots of experience raising and breeding organisms in captivity. For decades, we have raised animals in zoos, farms, breeding centers, and even our own houses. This experience has taught us that species differ in how they respond to captivity. Some species thrive—they are healthier, live longer, and have more offspring than wild-living individuals of the same species. Other species suffer terribly—they have shorter life expectancy, rarely reproduce, and even develop psychological disorders such as the repetitive swaying or pacing that is often observed in polar bears in zoos. Understanding how our candidate species for de-extinction is likely to fare in captivity will be critical to the success of our project.
When rearing genetically engineered animals, it will be important to keep in mind that some of the traits observed in these animals—both physical and behavioral traits—may be a consequence not of their edited genomes but of the pressures of life in captivity. One fascinating example of how different selection pressures in captivity can change the way an animal looks involves a wild population of Russian silver foxes. In 1959, Dmitry Belyaev, a Russian biologist who would later become director of the Institute of Cytology and Genetics, Russian Academy of Sciences, took 130 wild silver foxes and began breeding them on a farm near his institute in Novosibirsk. With each generation, he allowed only the foxes that appeared to be the tamest to breed. After only four generations, foxes began to wag their tails as their keepers approached. Over the course of just a few decades, his population of wild silver foxes was transformed into a population of animals that whined, wagged their tails, and jumped into the arms of and licked the hands of their keepers. The behavioral transformation was fascinating, but so was their physical transformation. The younger generations included animals with floppy ears, rolled or shortened tails, and coat coloration that is not seen in the wild but is reminiscent of other domestic animals.
Animals that are born and raised in captivity tend to look and act differently than their wild cousins. Physical differences appear in animals that are bred in captivity, including having shorter intestinal tracts and brains that are different sizes from those of wild-bred individuals. Traits that advertise sexuality also tend to be less noticeable among captive-bred individuals, which can influence their ability to find and compete for a mate in the wild. The behavioral differences between captive-bred and wild-bred individuals may be even more troubling. Animals in captivity don’t need to learn how to avoid predation, for example, and the absence of social conflict and the unnatural social structures that form in captivity can lead to changes in defensive or sexual behaviors. Without sufficient space or stimulation, some species become overwhelmed by stress in captivity and develop acute psychological disorders, such as wall licking by captive giraffes and self-mutilation by captive big cats and bears. Stress also affects the animals’ physiology, often reducing their fertility or stopping reproduction entirely.
Captive breeding also leads to unintended genetic changes, which may complicate the interpretation of genome-editing results. Without the need to find food, avoid predators, or fight off disease, selection pressures on animals living in captivity are relaxed. Instead of favoring traits for survival in the wild, captive breeding favors traits that increase reproductive success in captivity. This is not ideal when the goal is to release these populations into the wild.
Given the problems that so many species experience in captivity, it would be useful to have some way to predict a priori how a species might fare during this stage of de-extinction. It is tempting to guess how an extinct species might fare in captivity based on how well their living relatives fare. However, data from zoos and other captive-breeding facilities show that even close evolutionary relatives can differ considerably in their response to captivity. Among cetaceans, for example, captive Fraser’s dolphins and Dall’s porpoises damage their bodies by throwing themselves against the sides of enclosure pools and refuse to eat, while bottlenose dolphins and finless porpoises seem happy and playful in captivity, with reproductive rates that are sometimes higher than those observed in wild populations.
Those species that do fare well in captivity, from an animal welfare perspective at least, tend to be those that flourish in close proximity to humans. They include species that are sometimes characterized as “invasive,” such as rats and mice, and species that do well in urban environments. They are species that deal well with disturbance and are flexible in their reaction to predators and new resources. In a sense, species that do well in captivity are those that are not particularly likely to be extinct in the first place.
YET ANOTHER SET OF MAMMOTH CHALLENGES
Elephants, unfortunately, are among those species that do not fare well in captivity. Both African and Asian elephants live longer in the wild than they do in zoos. Elephants in zoos are prone to obesity, arthritis, and infections, particularly in their feet. Even worse, both living species of elephant struggle to reproduce in zoos. Their ovulation cycles become abnormal and unpredictable, and they have low fertility rates and high infant mortality rates. Many elephants living in zoos also show signs of psychological distress, including repetitive swaying behavior, hyper-aggression toward other elephants, and a propensity to kill their infants. These animals are provided with food, water, and medical care, and yet it seems clear that their most basic needs are not being met.
Elephants are known to be intelligent, social, wide-ranging animals and have needs that are very difficult to satisfy within the confines of most enclosures. There is little reason to suspect that the physiological and psychological needs of elephants whose genomes contain some small fraction of mammoth DNA will differ considerably from those of elephants whose genomes have not been edited. If elephants are going to be used in future de-extinction projects, sincere efforts will be necessary to improve the well-being of elephants in captivity and elephants released into the wild. This includes both careful consideration in the design of any enclosure in which these animals will breed and live, and the establishment of sufficiently large numbers of these animals in the wild to satisfy them socially and intellectually.
The challenges of captive breeding are likely to vary considerably among species. For example, species that migrate annually over long distances may be particularly unsuited to captive breeding, as sufficient space for this behavior will be exceedingly hard to replicate in a captive setting. If migratory paths are learned by interaction with a social group, how should scientists replicate the process by which these behaviors are learned?
Passenger pigeons were not migratory birds. However, they did fly long distances in order to find forests with sufficient numbers of fruiting trees to sustain their large flocks. Fledgling passenger pigeons learned this behavior by opportunistically joining flocks as they passed overhead. In his presentation at the TEDx event in March 2013, Ben Novak presented a plan to teach resurrected passenger pigeons how to find food. He proposed painting hundreds or thousands of homing pigeons so that they looked like passenger pigeons and training these painted homing pigeons to fly over the breeding colonies. These “surrogate flocks,” as he called them, would attract the attention of the fledglings, who would follow their instincts and join the flocks. The surrogate flocks would ferry the young passenger pigeons between feeding sites that Ben intended to set up across the northeastern United States. As the passenger pigeon population grew, Ben would gradually use fewer and fewer surrogate birds in his flock, until eventually only passenger pigeons remained, complete with behaviors taught to them by Ben via his trained, painted flock of homing pigeons.
The combination of captivity-induced stress, reproductive problems, genetic consequences of different selection pressures, and lack of appropriate social interactions in captivity perhaps explains why captive-breeding programs for conservation—those that aim to rear endangered species in captivity and eventually release them into the wild—have been so variable in their success. Some strategies to resurrect extinct behavioral traits in captivity, such as that proposed by Ben Novak to train passenger pigeons to find food, are so far-fetched that they might actually work. There is no doubt, however, that captive breeding will be another tall hurdle for de-extinction.
Yet, captive breeding may not be as high a hurdle as the step that would come next: releasing these genetically modified organisms into the wild and allowing them to fend for themselves.