Environment and conservation

The quiet application: reading, not writing

Most of what gene technology has actually changed in conservation involves no engineering at all. Environmental DNA sampling detects species from the traces they leave in water or soil, which lets a survey team confirm the presence of a rare amphibian or an invasive fish without catching anything. DNA barcoding identifies seized timber, fish fillets and wildlife products where morphology has been destroyed, which is how CITES enforcement cases are built.

Reference genomes underpin both. Large collaborative efforts such as the Vertebrate Genomes Project and the broader Earth BioGenome Project exist to build high-quality genomes across the tree of life, and their practical value to conservation is that population genomics — measuring inbreeding, historical population size and adaptive variation — needs a reference to measure against. Managing a small population sensibly is largely a question of knowing how much diversity is left and where it sits.

Engineered insects

The clearest deployments involve mosquitoes. Oxitec's engineered Aedes aegypti carry a self-limiting gene: released males mate with wild females, and female offspring do not survive to adulthood, so the local population falls and the introduced construct disappears from it over subsequent generations. Releases have been conducted in Brazil and, under regulatory authorisation, in parts of the United States. The approach is closer in spirit to the long-established sterile insect technique, which uses irradiation, than to anything self-propagating.

Target Malaria, a not-for-profit research consortium working with institutions in Burkina Faso, Uganda, Ghana and elsewhere, has taken a deliberately staged path towards Anopheles gambiae, the main malaria vector in much of sub-Saharan Africa: sterile male releases first, then self-limiting constructs, with gene drive as a later phase contingent on regulatory and community approval at each step. The sequencing is the point — it is a governance model as much as a research plan.

What a gene drive actually does

Ordinary inheritance gives a gene a coin-flip chance of being passed on. A homing gene drive carries the machinery to cut the partner chromosome and copy itself across, so nearly every offspring inherits it. That converts a small release into population-wide spread over generations, which is exactly why it is both attractive for disease control and alarming for anything that crosses a border.

Three limitations are well established. Resistance alleles arise when repair produces a sequence the drive can no longer recognise, and they are favoured by selection. Spread depends on population structure and mating behaviour, so models built on well-mixed populations flatter the technology. And containment is not a property you can add afterwards: designs that limit spread — split drives needing two components, or drives that run out of steam after a set number of generations — have to be built in from the start.

Trees, corals and islands

The American chestnut is the long-running plant case. A blight-tolerant line was developed carrying an oxalate oxidase gene from wheat, which detoxifies the acid the fungal pathogen uses to kill tissue; it has been through federal regulatory review in the United States and has divided the restoration community, partly on performance and partly on principle. Restoring a forest tree means releasing something intended to persist for centuries, and no trial period matches that timescale.

Coral work is mostly at the assisted-evolution and selective-breeding stage — heat-tolerant symbionts and parent selection rather than engineered corals. Proposals to use genetic tools against invasive rodents on islands, where eradication by trapping and toxins is the current practice, sit at the research and modelling stage and raise the sharpest containment question of all: an island is only a closed system until something gets on a boat.

Who decides

Deliberate release of a living modified organism falls under national biosafety law and, internationally, the Cartagena Protocol on Biosafety, which sets out advance informed agreement procedures for transboundary movement and the Biosafety Clearing-House through which decisions are notified. Parties to the Convention on Biological Diversity have repeatedly taken up synthetic biology and gene drives at conferences of the parties, and the IUCN has debated the role of synthetic biology in conservation at its congresses. Where a release affects communities directly, free, prior and informed consent is the framework most programmes now say they are working to.

The trade-off we keep coming back to

Conservation decisions are already made under uncertainty, and doing nothing is not neutral: habitat loss, invasive species and disease continue while the discussion runs. But an engineered organism released into an ecosystem is harder to recall than a pesticide and harder to monitor than a translocated animal. The honest position, as far as we can tell, is that self-limiting tools are defensible now with good monitoring, while self-propagating drives need containment designs, field-relevant modelling and community agreement before anyone should be comfortable with them.