An unmanned, AI-controlled, terraforming ship arrives at an earth-like planet (i.e in the goldilocks zone of its star, with a magnetosphere, an atmosphere, plenty of liquid water, earth-like gravity, etc.).

Despite this habitability, the planet has not developed any life at all.

The terraforming ship sets to work and succeeds in making the planet adequate for Earth-native species (altering the atmosphere composition for example).

Now it is ready to introduce earth species to the planet in order to prepare it for a colony ship that will arrive in a few hundred years time [edit: some answers suggest more time may be necessary to establish a functioning ecological pyramid, so specify more time if necessary].

What ecological pyramid will the ship introduce to create a viable, human-compatible ecosystem and food web?

I assume that:

  1. The terraforming ship will introduce the species in the temperate or Mediterranean-like zone of the planet, where humans can live comfortably.
  2. There's not enough time for species to evolve before the humans arrive, but the terraforming ship could potentially genetically-engineer some existing earth-species, if necessary.

I assume that the ecological pyramid will:

  1. Include bacteria, plants, herbivores, carnivores and predators.
  2. Create an ecosystem that can survive without human intervention, there being hundreds of years before the humans will arrive.
  3. Support human-consumable food of the sort earth-based farms produce e.g. wheat, rice, potatoes, vegetables, nuts, dairy foods, fruit, meat.
  4. Will be capable of supporting human useful animals like dogs, cats, sheep and horses, although the terraforming ship may wait to introduce them until the humans arrive if that's necessary.
  5. May include birds, fish, spiders, insects and other species, but only if they are needed as part of the pyramid. It will not include any 'unnecessary' species. For example, there's no need for kangaroos or penguins.

I'm happy to amend or clarify the question if it's unclear or I've made invalid assumptions.

  • 2
    $\begingroup$ Biosphere 2. $\endgroup$
    – AlexP
    Commented Nov 28, 2020 at 16:07
  • 3
    $\begingroup$ Whats the composition if the atmosphere? If there are no pre existing, carbon based, native lifeforms & no photosynthesis then there will not be enough 'free' oxygen in the water or atmosphere to support aerobic respiration (animal life). That being the case your colonists either delay their arrival by centuries while the oxygen cycle is established using algae, mosses and lichens etc or else accept the fact they'll be living in dome colonies and moving around on the surface with respirators for generations to come while it does get established. $\endgroup$
    – Mon
    Commented Nov 29, 2020 at 0:34
  • $\begingroup$ If there is no current life, then 99.9% of the living things that we know on earth could not survive there. For instance, if there is no life there, then there is no soil. Do you know how long it takes to make enough soil for plants, starting from just anaerobic bacteria? I don’t either, but I’m pretty sure it has a lot of zeroes at the end. $\endgroup$ Commented Nov 29, 2020 at 4:15
  • $\begingroup$ No soil yes but there will be regolith i.e. eroded & pulverized rock granules. Which is basically all soil is if you take out the carbon & organic material. So eventually they can make soil - once you have the oxygen cycle issue solved. $\endgroup$
    – Mon
    Commented Nov 29, 2020 at 6:00
  • 2
    $\begingroup$ In one of Kim Stanley Robinson's Red Green Blue Mars books one of the characters attempts soil creation from scratch - with much detail provided. The books do this also planet wide but with less detail. If you've not read them you should - they provide a superb world building in all areas view. $\endgroup$ Commented Nov 29, 2020 at 6:51

5 Answers 5


Spirulina. That's it.



These blue green algae will happily take the ocean over and probably large swaths of the land too. They once ruled the earth and they still do very well. They are nutritious food for man and beast. Once the colonists show up they can harvest huge rafts of the stuff and use it as soil amendments, possibly with a soil starter culture.

It is conceivable that this is all they would need, though probably not all they would want. And it makes for a good fiction - looking out over the gooey green ocean, thinking about a corn tortilla...

  • 2
    $\begingroup$ Cyanobacteria is the reason we have oxygen in our atmosphere and gas in our cars. But they lack some vitamins (notably B12) for our survival. I say algae first, then fish after a while. $\endgroup$ Commented Nov 29, 2020 at 18:43

Lavender, Bees, Evergreen trees, Phytoplankton, Herrings, and some bacteria. That's it.

Everything else can wait for just before humans arrive. This is the absolute minimum you'll need.

  • Lavender is high in nectar and flowers all year round in many climates. It's ideal for bees to gather nectar from and to maintain a bee population.
  • Bees you need to fertilise your crops. Let them be (ha!) for the moment, and let them grow in population.
    • I don't think you'll need spiders to control the bee population (they'll be limited by the lavender food supply), but that's an optional extra that may be needed.
  • Evergreen trees. Probably pine. Makes a nice construction material for the humans when they get there.
  • Fill the oceans with Phytoplankton. This consumes CO2 and generates oxygen in the water, allowing fish to eventually survive, this eventually gets in the air and makes the air nice and breathable for humans.
    • There are lots of types. Your terraformers should experiment and determine which one is optimal for the exact light / gas / mix of your planet.
  • Nitrosomonas or Nitrobacter to make nitrates, or convert nitrites to nitrates. There are many choices, just pick the best one which works experimentally on the exact atmosphere you have. (Is there ammonia gas or triple bonded nitrogen?) Examples are Azotobacter, Bacillus, Clostridium, and Klebsiella.
  • Herring fish (these eat plankton and are good for humans). Without predators their population will be controlled by the plankton supply - they eat too much, they run out of either oxygenated water or food. These also make sure the plankton levels don't get too high - as too little CO2 and your planet will cool.

Your terraforming bots should add phosphorus and potassium to the soil. Your settlers will arrive with fertilisers to help with creating good farmland. Your terraforming bots should also use geological processes to generate some initial oxygen if possible.

Everything else should be grown in farms as much as possible.

  • 1
    $\begingroup$ "Plankton" means "protists or small animals or plants which live in water and float with the current". There are uncountable species of protists, animals and plants which are part of the plankton. ("Plankton", from Greek planktos = drifting, is the opposite of "nekton", from Greek nêktos = swimming, meaning animals or plants which live in water and propel themselves actively.) $\endgroup$
    – AlexP
    Commented Nov 28, 2020 at 16:04
  • 1
    $\begingroup$ "Phytoplankton" is a general term covering uncountable species of bacteria, diatoms, dinoflagellates, algae and so on. Do you need all of them, most of which are yet unknown to science? (And a planet with no life does not have any "soil", by definition. Soil is a product of life. On planets without life you can only have regolith.) $\endgroup$
    – AlexP
    Commented Nov 28, 2020 at 16:27
  • 7
    $\begingroup$ I think you are underestimating the complexity of fertile soil (as opposed to sand). $\endgroup$
    – o.m.
    Commented Nov 28, 2020 at 16:27
  • 1
    $\begingroup$ What would happen to dead trees, bees, humans and herring? Chemical processes of decomposition might not be sufficient to remove them fast enough (depending on circumstances) in which case they would build up. I suggest some form of fungus might be useful $\endgroup$
    – Slarty
    Commented Nov 28, 2020 at 22:48
  • 1
    $\begingroup$ If there’s no life on this planet, then there’s no soil either. Most plants need soil. $\endgroup$ Commented Nov 29, 2020 at 4:09

...and then a miracle happens...

I think you probably can't have a fully-functional and evolved ecosystem of an entire planet in a few hundred years. Sorry. It would likely take that long to build an underlying ecology of microorganisms, assuming they are all evolved to survive the local conditions.

An ecosystem consists of so many moving parts that all need to be fully functional. It's not really sufficient to terraform a small region when you have an entire planet to deal with. What you can have is a widely distributed set of microorganisms (like blue-green algae) that have set the preconditions for an ecosystem. I doubt there would be time to fully establish a functioning oxygen atmosphere (the process requires the free oxygen to react with vast amounts of things like free iron in the environment, a process that took millions of years on Earth).

Given hyper-efficient super organisms, perfect starting conditions, abundant raw organic materials (yes, there were organic materials before life was established on Earth) and a goldilocks planet, I'm going to say you can have an underlying microorganism population primed for more advanced organisms (still pretty simple, mostly plants in the ocean) in MAYBE 1000 years. Stuff has to grow, and even your terraforming ship would need raw materials to grow organisms, then dump them into the environment. There would likely be at least ONE event where some critical organism unexpectedly fails, setting the whole program back and requiring creative substitution.

Your ship will need to be pretty smart, and even then biology isn't like physics. While it does follow chemistry and physics, it's a complex and unstable system of moving parts. To establish a functioning eukaryotic multicellular biome, you'll probably need your colonists to arrive with their squishy brains and extemporize the ecosystems together.

That being said, once you achieve a stable, functioning ocean ecology (involving some fish, sponges and other filter feeders, etc) the problem of adding additional biomes becomes simpler. Once radiation is under control, plants can form the basis of ecologies and spread on the land. In a lot of ways, the whole process needs to recapitulate the evolution of life on Earth. The early plants will need to be simple and self-sufficient, and later you can get fancy, adding insects and other pollinating organisms, seed carriers, etc.

If you want to establish an ecosystem in domes, the problem becomes simpler. These could be the 'seeds' for later biomes on the surface. I would study the attempts to have self-sustaining biospheres in buildings that have been made. So far, they've all failed. That doesn't mean they won't eventually succeed, but it just proves the process is a lot more complicated than we'd like it to be.

  • $\begingroup$ Thank you for your answer, which is very interesting and makes some excellent points. However, it seems to me more like a comment on the subsidiary specification that it's "a few hundred years" until the colonists arrive, rather than a response to the main question of what a minimalist ecosystem might consist of. I can edit the question to remove the unrealistic time-frame, if that helps. $\endgroup$
    – Shimbo
    Commented Nov 28, 2020 at 18:02
  • $\begingroup$ @Shimbo It's really more about the fact that an ecosystem can't really be narrowed down to a minimalist subset. Every attempt to do so has ended in failure. A planet has to go through a distinct set of transformations to become a place friendly for life. The more minimalist your ecosystem, the more vulnerable to any kind of even minor shift. It becomes a house of cards, waiting for any sudden shift to send the whole thing crashing down. And the timeline is unrealistic. $\endgroup$
    – DWKraus
    Commented Nov 28, 2020 at 20:05
  • $\begingroup$ Just creating enough soil for plants to live would probably take at least a million years. $\endgroup$ Commented Nov 29, 2020 at 4:10
  • 1
    $\begingroup$ @RBarryYoung True, although that's actually a little more nuanced than you might think. You might be able to dredge sediment from the ocean and river beds, then enrich it with organics, minerals, and microorganisms. It can approximate soil. I've never heard of anyone trying it on a large scale. Some plants need surprisingly little soil (essentially none, by our standards), especially simple ones (the starter organisms). $\endgroup$
    – DWKraus
    Commented Nov 29, 2020 at 6:22
  • $\begingroup$ But how are you going to do that for an entire world? The OPs assumption seems to be to get the organism together, and then everything can live together, or at worst you stage the organisms so the earlier organisms can do the transformative work for the later ones. But soil alone takes millions of years, and that’s just one of the transformations that have to take place. Besides dry sediments (soil, etc.) there’s also oceanic sediments, chemistry, and particulates, fresh water contents, atmospheric contents (oxygen, etc.), and stabilization of oceanic currents and climate patterns... $\endgroup$ Commented Nov 29, 2020 at 15:42


There are a bunch of things that need to happen to have modern life happy on a planet, like the oxygen apocalypse. This happening "naturally" will take a long time.

Have your terraforming ship land robots that build huge solar power arrays, that in turn release massive amounts of oxygen into the atmosphere. It can convert CO2, H2O or even Oxygen trapped in Iron or similar materials into free oxygen.

You'll probably end up covering much of the planet in solar panels along the way, and it will still take a long long time. But your machines will be happy with a toxic-to-modern-life environment.

As you do this, start growing life forms that can survive in the low-oxygen environment and help the atmosphere along. Any kind of photosynthesis will do. You'll want machines able to "print" various kinds of microbes and widely seed them to speed this process up, and libraries of such microbes you use in each phase of the process.

Now, before you start any of this, your ship will have visited the asteroid belt. As an interstellar ship, it is going to be tiny, as energy budgets needed to travel between stars in "reasonable" periods of time are ridiculous. So it finds a broken protoplanet core like Davida and builds the rest of the resources it needs in the vacuum of space.

This process will continue as the planet is seeded. Orbital solar collectors will beam energy down to the planet, reducing the amount of infrastructure you have planet-side. If the planet is short water, missions to the ice zone will land comets on the planet (note this could add millions of years to the time required to terraform; moving planet-scale amounts of water to a planet is a high-energy act).

Or, if you are crazy, you could even drop multiple beanstalks from orbit, allowing for more efficient transport of water to the surface (reduce the time you spend boiling entire planet's oceans along the way).

The dirt of Earth is the result of many many millions of years of work as well. So mass producing machines that mechanically grind up soil and have internal bioreactors to prepare might be needed.

This process is going to be high energy, turning the planet into a hothouse, as you try to do in a mere 100s of years what usually takes millions. Cooling the planet may require extreme measures; blocking the sun with giant shades would be one step, but also using laser heat-pumps or increasing the radiant surface area of the planet with the beanstalks or more insane mechanisms.

Fortunately, long before you have a habitable planet, you'll be able to manufacture space stations, or bases buried under the surface of a moon (radiation shielding), supplied by stellar orbital asteroid mining.

Humans could arrive (or more realistically, be printed) long before you the planet is done. And in the relatively controlled environment of tunnels under a cold moon, you can gradually create more habitable terrain without having to finish an entire planetary biosphere.


I think you need to make it a two stage process;

This is because you state in your original scenario that 'the planet has not developed any life at all'. And this is key problem because it means there will be little if any free oxygen in the atmosphere and no Oxygen Cycle. Without some form of pre-existing photosynthesis or a similar chemical process releasing oxygen into the atmosphere constantly any free oxygen in the primeval atmosphere will have very early been oxidized and bound into the crust. That being the case.

Step one; An automated probe is sent ahead of the colonists (or perhaps the one that found the planet originally can do this). The probe carries a payload consisting of a wide spectrum of algae species selected and/or engineered for specific sub environments (marine, fresh water, soil/rock, alkaline, volcanic arctic/tropical etc . In short the full range of environments where algae thrives on Earth today. It also carries a range of nitrogen fixing bacteria and simple plants and communal algae for follow up drops.

The probe then seeds the planet, matching 'drop' locations with those species best suited to surviving in that local environment. After that it just sits back and sends back regualr reports on oxygen levels in the atmosphere and later seeds the local regolith with nitrogen fixing bacteria and plants.

Step two; Your scientists/colonists are sent once oxygen levels start to reach something approaching breathable levels and can then start planning/constructing proper biomes using species they bring with them.

How long between Step 1 and Step 2? I have no idea, perhaps an expert could tell you but I suspect it would be surprisingly rapid.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .