One candidate is gene drive, a technology capable of altering the genomes of wild populations. Professor John Knight writes in support (ODT 10.10.16), while Colin Campbell-Hunt urges caution (ODT 11.10.16).
As a scientist working in the field, I agree that a biological solution may be required, but only if it cannot spread far beyond the site of release. Attempting to use standard gene drive designs for conservation would be extremely unwise.
Harnessing gene drive only became possible with the development of CRISPR-based genome editing a few years ago. Encode a desired change along with the components of the CRISPR system, and it will cut and replace the original sequence with the new version over subsequent generations. Think of it as a “find and replace” for the entire population. As that metaphor implies, such drive systems are highly invasive: we refer to them as 'global' because they are likely to spread to every population of the target species.
In theory, gene drive systems can suppress or locally eliminate populations, with obvious implications for predator removal. Ensuring that females who inherit the drive system from both parents are born infertile is more technically feasible than ensuring that all offspring are fertile males, but either approach can work. Unlike traps and poisons, these methods would not cause any animals to suffer. But as my colleagues and I emphasized when we first described CRISPR-based gene drive in 2014, tackling rats with this strategy would necessarily be a worldwide endeavour.
Should New Zealanders release rats carrying a global suppression drive, the construct would almost certainly hitch a ride on ships or planes to other continents before it eliminated the local population and extinguished itself. If the gene drive rats did not manage to do this on their own – and there are few reasons to think they would not – it's safe to assume that someone would move them deliberately. After all, the species imposes considerable economic costs on many industries. The illegal release of rabbit calicivirus in New Zealand provides an instructive example.
Unless international spread is the explicit goal, it is never safe to release a global drive system in any area where the target organism is present. No matter the physical containment procedures used, it would be reckless to build a global drive system anywhere within the endemic region because a single accidental escape could eventually affect everyone who shares an ecosystem with that species. For rats and mice, that's most people on Earth.
To put it in the starkest of terms, making a global drive system is equivalent to creating a new highly invasive 'species' anticipated to spread around the world. One would think we have enough of those.
Some might say this is precisely why we should attempt to eliminate rats worldwide, at least outside of pet stores. Perhaps that would be a good idea, but doing so without the permission of every other country would be highly irresponsible. Because gene drive is arguably the technology most likely to help eradicate human diseases such as malaria and schistosomiasis, it would be a profound tragedy if New Zealanders – or anyone else – inadvertently caused an international incident and consequent loss of public confidence in scientists and governance. Going by past examples of accidents and misconduct in diverse fields of science, any unauthorized release of a gene drive system would quite likely delay beneficial applications by a decade or more. For malaria alone, the cost could be millions of unnecessary deaths.
Global drive systems should only be built to combat true plagues for which we have few other countermeasures, and then only in laboratories located far from existing populations of the target species. For all other candidate applications – including predator removal in New Zealand – we should focus on local drive systems that cannot spread indefinitely.
For example, my laboratory is currently developing “daisy drive” systems in which the drive components are linked in a daisy chain: D drives C drives B drives A. Because D can't drive, it will be lost in half of offspring, which causes C to stop driving and be lost, and so on. The effect is to make the links of a daisy drive function as genetic fuel: they're successively lost over generations until the drive runs out and stops. Our models suggest that releasing only a few such organisms can eventually affect a fairly large population, so releasing enough animals for local removal would cause little additional ecological damage.
While conceptually promising, there is much work to be done before we get any local drive system working in rats, let alone stoats and possums, and then even more to ensure that they will behave as intended. Humanity has no experience engineering systems anticipated to evolve outside of our control, so we will need rigorous safety testing in the laboratory before considering field trials.
While respectful of the daily toll that invasive predators inflict on native species, I suggest that interested communities and conservationists avoid pressing for global CRISPR gene drives in favor of applying existing measures and helping to guide the development of local drive approaches. The probable cost of impatience is simply too high.
Even so, proponents of genetic solutions are right to start the conversation now. Any technology capable of eliminating mammalian predators from New Zealand will by definition alter the shared environment. That means that all proposals and research should be open from the earliest stages, with researchers and supporters actively inviting concerns and criticism from people with reservations.
There's a running joke that “CRISPR time” runs at least 10 times faster than a normal scientific field. Even if daisy drive and variants don't work out, it's quite likely that a superior strategy will be invented in the next several years. Given that New Zealand's target date for mammalian predator removal is 2050, the requisite technologies should be available well in advance. Are New Zealand communities prepared to guide the development and oversee the testing of these systems? I certainly hope so, because scientists shouldn't be doing this on their own.
- Kevin M. Esvelt is an assistant professor and leader of the Sculpting Evolution Group at the MIT Media Lab.
