There are two possible scenarios for the future of humankind in the Anthropocene:
Scenario 1: we catastrophically over-reach the earth's carrying capacity, resulting in an uncontrolled crash back to a pre-industrial state (from which we will probably never fully recover given that all the easily accessible fossil fuels are now used up);
Scenario 2: we manage the coming demographic transition in a sustainable way, with population (and resource utilisation) peaking sometime mid-Century followed by a gradual decline in both population and resource use, and with continued increase in our technological capacity.
In either scenario, the Anthropocene will inevitably comprise a mass extinction event. In the first scenario, the living world will come through the mass extinction in much the same way as it has in the past, with diversity gradually rebuilding over the next several million years.
In the second scenario, we have a more interesting (and optimistic) scenario. After the demographic transition, we will be able to gradually reduce our ecological footprint, and will be able to embark on a phase of "rewilding" the planet - we contract the amount of land we need for food production as we farm more intensively and technologically for a smaller population, which makes land available for restoring wild spaces and the ecological services these provide.
The problem is, we will have lost a significant amount of biodiversity by the time we can start the rewilding.
So here's the proposal:
By 2028 we will have collected and stored tissue samples from all Australasian biota in a genomic ark
Despite the inevitable mass extinction, we act now to store as much genetic diversity as possible in the form of DNA-stable tissue samples. Once the rewilding starts, we will almost undoubtedly have the technological capacity to recreate species from DNA samples, using CRISPR/Cas9 or other gene editing technologies. The limiting factor will be having the DNA material to do this.
So, by 2028 we will have an Australasian genomic ark, the main purpose of which (though not the only purpose) is to store stable genomic samples of as many organisms as possible from our region. This can be done in a variety of ways, e.g. dried leaf samples for plants and tissue samples for animals, environmental samples for fungi, microbes etc.
An interesting issue is that we need to try to sample as much biodiversity as possible including currently unknown taxa. The taxonomy can come later (perhaps even after the rewilding) - we don't need to have the taxonomy all worked out before we sample. We should aim to store as many tissue samples as possible, whether the know what the organism is or not. This has interesting implications for the sampling strategy we would employ for this.
Such an ark could become an absolutely invaluable resource in the future (if we can avoid scenario 1).